INDEX

Explanations

the word "compare" in various contexts

oai_token-act-pair · gpt-3.5-turbo

instances of the word "compare" 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 @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

vous

-0.89

bott

-0.69

gren

-0.69

ktop

-0.69

der

-0.64

Cmd

-0.63

noon

-0.63

Sky

-0.62

oÄŁ

-0.62

fly

-0.61

POSITIVE LOGITS

 favorably

0.89

 apples

0.86

 Compare

0.85

 comparisons

0.80

isons

0.80

uations

0.76

rals

0.75

onga

0.74

 compare

0.74

 objectively

0.73

Activations Density 0.018%

the word "compare" in various contexts

instances of the word "compare" in various contexts

No Comments

No Known Activations