INDEX

Explanations

words related to the action of retreating or withdrawing

oai_token-act-pair · gpt-3.5-turbo

the word "retreat" and its variations, indicating themes of withdrawal or retreating

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

IFrame

Link

Negative Logits

ym

-0.73

 purse

-0.68

GB

-0.66

tha

-0.63

Lev

-0.63

 Cosponsors

-0.63

colo

-0.61

oven

-0.60

 Parenthood

-0.60

nea

-0.60

POSITIVE LOGITS

sburg

0.96

 retreat

0.88

alus

0.78

oleon

0.75

dinand

0.75

icter

0.73

EStream

0.73

aroo

0.72

llan

0.70

eatures

0.70

Activations Density 0.012%

words related to the action of retreating or withdrawing

the word "retreat" and its variations, indicating themes of withdrawal or retreating

No Known Activations