INDEX

Explanations

words related to fairness and bias

oai_token-act-pair · gpt-3.5-turbo

words related to impartiality and bias

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

Phones

-0.78

fts

-0.76

 Danger

-0.73

Memory

-0.72

Pause

-0.71

 Nightmares

-0.71

Fly

-0.70

Requires

-0.70

phrase

-0.69

waves

-0.68

POSITIVE LOGITS

 impartial

1.33

 unbiased

1.26

 observer

1.07

 biased

0.89

 observers

0.88

 neutral

0.86

 arbit

0.85

 overse

0.79

 appra

0.78

 opinions

0.78

Activations Density 0.029%

words related to fairness and bias

words related to impartiality and bias

No Comments

No Known Activations