INDEX

Explanations

phrases related to news interviews and press conferences

oai_token-act-pair · gpt-3.5-turbo

punctuation marks, specifically commas, in the 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

SL

-0.72

alogue

-0.71

orest

-0.70

TRY

-0.68

isSpecialOrderable

-0.68

role

-0.67

Cover

-0.67

rities

-0.66

cos

-0.65

ãħĭãħĭ

-0.65

POSITIVE LOGITS

 however

0.88

 meanwhile

0.79

 Carlo

0.78

 excerpts

0.76

 Berm

0.76

 though

0.75

 dated

0.75

 citing

0.74

 noting

0.73

 titled

0.71

Activations Density 0.202%

phrases related to news interviews and press conferences

punctuation marks, specifically commas, in the text

No Comments

No Known Activations