|
pulsatrix 1.1.0
|
A trainable edit of the residual stream after a transformer block (TRN-9: LoReFT), run by CausalLM::set_intervention(). Maps hidden states (N, L, hidden_size) to the same shape and, as a Module, trains and explains with the model.
More...
#include <causal_lm.hpp>


Public Member Functions | |
| virtual Tensor | forward_at (const Tensor &hidden, int64_t start, int64_t prompt_length)=0 |
The edit for hidden states covering positions [start, start + L) of a sequence whose prompt is prompt_length positions long. forward(x) is forward_at(x, 0, L). Cached generation calls it once on the prompt and then once per new token, each a later start. | |
Public Member Functions inherited from pulsatrix::Module | |
| virtual | ~Module ()=default |
| Tensor | forward (const Tensor &input) |
| Runs this module's forward computation. | |
| virtual std::optional< DeviceType > | compute_device () const |
| The device this module computes on, so forward() can reject an input on another device before any kernel sees it (roadmap FND-8). | |
| std::pair< Tensor, NodeId > | forward_traced (const Tensor &input, NodeId input_node, ComputationGraph &graph, Autograd &autograd) |
| Runs forward() while also registering a ComputationGraph node (tagged with this module's op_type(), parented to input_node) and wiring an Autograd backward function that reuses this module's own backward() – the opt-in traced/explainable path, per Phase 2 Mission 0. | |
| virtual Tensor | propagate_relevance (const Tensor &relevance_out, const LRPRuleConfig &config)=0 |
| Computes this module's contribution to LRP relevance propagation. | |
| virtual bool | supports_lrp_rule (LRPRule rule) const |
Whether propagate_relevance() implements rule (no silent fallback: callers such as ExplainerContext::relevance_pass() throw rather than run a module on a rule it does not implement). | |
| virtual Tensor | backward (const Tensor &grad_output)=0 |
| Computes the gradient w.r.t. this module's input, given the gradient w.r.t. its output. Must be called after forward() – uses state cached from that call. | |
| virtual OpType | op_type () const =0 |
| This module's operation-category tag, for ComputationGraph node tagging. | |
| virtual std::vector< NamedParamRef > | named_parameters () |
| This module's trainable parameters, each with its hierarchical name – the one place a module declares its parameters (roadmap FND-1). | |
| virtual std::vector< NamedBufferRef > | named_buffers () |
| This module's buffers: state a checkpoint must save that no optimizer updates. | |
| virtual std::vector< NamedModuleRef > | named_children () |
| This module's direct sub-modules, each under the prefix its parameters carry in named_parameters(), in the same order (TRN-7: how ApplyLora() finds a model's Linear layers). Default: none. NamedModules() walks the whole tree. | |
| virtual std::vector< ParamRef > | parameters () |
| This module's trainable parameters and their gradients, for an optimizer to update uniformly across module types. | |
| void | set_requires_grad (bool requires_grad, const std::string &prefix="") |
Freezes (false) or unfreezes (true) parameters by name (roadmap FND-2). | |
| virtual void | set_training (bool training) |
| Sets this module's training/eval mode. Defaults to training (matches every mainstream framework's Module default). | |
| bool | is_training () const |
| Whether this module is currently in training mode. | |
| virtual void | release_activations () |
| Frees the tensors forward() kept for backward() and the relevance passes; containers release their children too. Inference that only needs the output calls it after each forward, so a deep model holds one layer's activations at a time (a 1024-token ESM-2 650M pass otherwise keeps over 10 GB of them). | |
Protected Member Functions | |
| Tensor | forward_impl (const Tensor &hidden) override |
| The actual forward computation. Called by forward() after precondition checks. | |
A trainable edit of the residual stream after a transformer block (TRN-9: LoReFT), run by CausalLM::set_intervention(). Maps hidden states (N, L, hidden_size) to the same shape and, as a Module, trains and explains with the model.
|
pure virtual |
The edit for hidden states covering positions [start, start + L) of a sequence whose prompt is prompt_length positions long. forward(x) is forward_at(x, 0, L). Cached generation calls it once on the prompt and then once per new token, each a later start.
Implemented in pulsatrix::HiddenStateTap, and pulsatrix::LoReFTIntervention.
|
inlineoverrideprotectedvirtual |
The actual forward computation. Called by forward() after precondition checks.
Implements pulsatrix::Module.