INDEX

Explanations

mentions of specific names or terms related to Tol

oai_token-act-pair · gpt-3.5-turbo

references to specific names and the theme of vilification

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

 depreciation

-0.66

ILY

-0.64

 showers

-0.64

 shower

-0.63

boards

-0.63

board

-0.62

manship

-0.62

 stump

-0.62

eering

-0.61

eding

-0.61

POSITIVE LOGITS

 Nadu

0.82

sie

0.82

edo

0.82

ove

0.81

achev

0.80

entially

0.79

ibr

0.78

irtual

0.77

ieve

0.73

ima

0.73

Activations Density 0.027%

mentions of specific names or terms related to Tol

references to specific names and the theme of vilification

No Comments

No Known Activations