INDEX

Explanations

specific mentions of the word "format"

oai_token-act-pair · gpt-3.5-turbo

mentions of different formats or structures

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

roma

-0.84

doms

-0.82

guard

-0.73

arma

-0.73

nee

-0.72

hiro

-0.72

ĺħ

-0.70

ortium

-0.69

minent

-0.69

 brow

-0.68

POSITIVE LOGITS

ters

1.09

ting

0.88

atted

0.85

 format

0.84

tered

0.80

aldehyde

0.79

 Format

0.78

 formats

0.77

etter

0.77

tering

0.77

Activations Density 0.027%

specific mentions of the word "format"

mentions of different formats or structures

No Comments

No Known Activations