INDEX

Explanations

titles or headings containing special characters and numbers in a specific format

oai_token-act-pair · gpt-3.5-turbo

the symbol 'Ŀ' or similar variations indicating a specific coding or character-related focus

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 ende

-0.81

 ponder

-0.80

 psychiat

-0.75

 unwanted

-0.73

 contrace

-0.73

 incorpor

-0.72

 unconscious

-0.72

 recip

-0.72

 sacrific

-0.71

 unintended

-0.70

POSITIVE LOGITS

Â¯

0.95

ï¸

0.86

é¾į

0.85

LED

0.82

log

0.80

xxx

0.78

times

0.78

Â°

0.78

Hall

0.78

DAY

0.78

Activations Density 0.196%

titles or headings containing special characters and numbers in a specific format

the symbol 'Ŀ' or similar variations indicating a specific coding or character-related focus

No Comments

No Known Activations