INDEX

Explanations

words related to problem-solving or decision-making

oai_token-act-pair · gpt-3.5-turbo

action-related words that indicate processing, voting, or decision-making

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ancies

-0.61

 Carlo

-0.60

asio

-0.60

enegger

-0.59

lesi

-0.58

 Pirates

-0.58

anu

-0.57

 Corvette

-0.57

acion

-0.56

utra

-0.56

POSITIVE LOGITS

 upon

0.90

awa

0.83

 down

0.79

chy

0.79

 anew

0.78

up

0.77

Ô

0.76

 into

0.76

 merciless

0.75

ãĤ¡

0.75

Activations Density 0.172%

words related to problem-solving or decision-making

action-related words that indicate processing, voting, or decision-making

No Comments

No Known Activations