INDEX

Explanations

instances of emotional expressions or reactions

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

 avoidance

-0.76

Suk

-0.63

 pressures

-0.61

 Companion

-0.61

 Freed

-0.58

 Drawn

-0.58

 Mole

-0.58

Hug

-0.57

 Blend

-0.56

ABE

-0.56

POSITIVE LOGITS

ever

0.92

ttle

0.89

ï¸ı

0.87

lee

0.86

ggle

0.84

athom

0.83

reci

0.81

conom

0.80

etary

0.78

ufact

0.78

Activations Density 0.139%

instances of emotional expressions or reactions

No Comments

No Known Activations