INDEX

Explanations

words related to accountability and responsibility

oai_token-act-pair · gpt-3.5-turbo

variations of the word "hold."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

endix

-0.80

ãĤ¡

-0.73

ghan

-0.71

ibel

-0.70

shire

-0.69

lez

-0.69

atis

-0.69

————

-0.68

é¾įå

-0.67

naire

-0.66

POSITIVE LOGITS

 sway

1.07

 onto

1.06

 accountable

1.06

 hearings

0.89

hold

0.86

overs

0.85

 hostage

0.85

erness

0.85

 conferences

0.82

holders

0.81

Activations Density 0.039%

words related to accountability and responsibility

variations of the word "hold."

No Comments

No Known Activations