INDEX

Explanations

the word "for" in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

orthy

-0.81

rete

-0.79

oS

-0.79

dar

-0.78

zu

-0.74

rets

-0.71

yn

-0.69

irl

-0.68

shi

-0.67

osterone

-0.66

POSITIVE LOGITS

etheless

1.01

 nonetheless

0.99

 reality

0.86

 nevertheless

0.78

 sheer

0.77

 hindsight

0.73

 actual

0.71

 Garg

0.70

 retrospect

0.67

 Builder

0.63

Activations Density 0.153%

the word "for" in various contexts

No Comments

No Known Activations