INDEX

Explanations

phrases indicating a comparison, often suggesting a superior or dominant position

oai_token-act-pair · gpt-3.5-turbo

the word "out" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

å£«

-0.72

otaur

-0.65

 Municip

-0.63

ahime

-0.63

 Symbol

-0.61

 Brach

-0.60

idate

-0.56

vernment

-0.54

 Minotaur

-0.54

 Decay

-0.53

POSITIVE LOGITS

fitted

1.12

 loud

0.95

number

0.93

posts

0.93

fitting

0.92

lasting

0.88

stretched

0.86

doors

0.86

last

0.86

range

0.83

Activations Density 0.066%

phrases indicating a comparison, often suggesting a superior or dominant position

the word "out" in various contexts

No Comments

No Known Activations