INDEX

Explanations

phrases indicating progress or improvement

oai_token-act-pair · gpt-3.5-turbo

phrases indicating progress or distance traveled toward a goal

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.7.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

urated

-0.80

iries

-0.70

ividual

-0.63

uala

-0.63

ulhu

-0.63

icist

-0.60

rones

-0.59

zinski

-0.58

 pairs

-0.58

iasco

-0.58

POSITIVE LOGITS

 toward

0.91

 towards

0.87

WARD

0.74

fare

0.69

 Towards

0.68

lier

0.66

 Sabha

0.64

 Drawn

0.63

finder

0.63

 separating

0.62

Activations Density 0.033%

phrases indicating progress or improvement

phrases indicating progress or distance traveled toward a goal

No Comments

No Known Activations