INDEX

Explanations

expressions indicating a judgment or evaluation of a situation

oai_token-act-pair · gpt-3.5-turbo

the word "just" and its variations used 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 @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 seiz

-0.88

 sidx

-0.77

rimination

-0.70

pora

-0.70

SEE

-0.70

 Also

-0.69

ixel

-0.68

iere

-0.67

itsch

-0.66

luaj

-0.66

POSITIVE LOGITS

ifiable

1.42

 plain

1.03

ifications

0.97

 icing

0.87

ified

0.87

 another

0.86

 coincidence

0.83

 kidding

0.83

 scratching

0.82

another

0.80

Activations Density 0.056%

expressions indicating a judgment or evaluation of a situation

the word "just" and its variations used in various contexts

No Comments

No Known Activations