INDEX

Explanations

words related to negative or critical connotations

oai_token-act-pair · gpt-3.5-turbo

empty or blank segments of text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 Sunshine

-0.56

©¶æ

-0.54

 Comet

-0.53

-+-+

-0.50

-0.49

 precinct

-0.49

roller

-0.48

Newsletter

-0.48

ļé

-0.48

Ĭ±

-0.48

POSITIVE LOGITS

wise

0.76

etheless

0.71

pecially

0.61

cause

0.57

tis

0.57

fortunately

0.56

lly

0.54

wat

0.53

:=

0.53

 versa

0.52

Activations Density 0.562%

words related to negative or critical connotations

empty or blank segments of text

No Comments

No Known Activations