INDEX

Explanations

contractions of words with negation

oai_token-act-pair · gpt-3.5-turbo

negations or expressions of uncertainty

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 Parenthood

-0.68

 Companies

-0.66

 Compass

-0.66

 Passage

-0.66

 Conversation

-0.65

 Handbook

-0.65

ounters

-0.62

arten

-0.61

 Personality

-0.60

artney

-0.60

POSITIVE LOGITS

 necessarily

1.04

 hesitate

0.92

oooooooo

0.91

 kidding

0.89

 gonna

0.86

 ashamed

0.82

 myself

0.82

 remem

0.82

 recommend

0.82

 believe

0.82

Activations Density 0.147%

contractions of words with negation

negations or expressions of uncertainty

No Comments

No Known Activations