INDEX

Explanations

words related to professional, organizational, and governmental entities

oai_token-act-pair · gpt-3.5-turbo

terms related to social surveys and general statistics

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

 cones

-0.76

Ly

-0.65

nen

-0.63

NZ

-0.61

ĵĺ

-0.60

 Slate

-0.58

 poles

-0.57

 havens

-0.57

 Rowling

-0.56

 generations

-0.56

POSITIVE LOGITS

iple

0.77

 guiActiveUnfocused

0.74

++)

0.71

 Reserved

0.68

utenant

0.66

onnaissance

0.65

emetery

0.65

gage

0.64

å°Ĩ

0.63

istries

0.62

Activations Density 0.143%

words related to professional, organizational, and governmental entities

terms related to social surveys and general statistics

No Comments

No Known Activations