INDEX

Explanations

words related to challenges or problematic situations

oai_token-act-pair · gpt-3.5-turbo

words related to conflict and contention

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

rontal

-0.57

pread

-0.56

deen

-0.51

urred

-0.51

 BaseType

-0.49

hower

-0.49

 Referred

-0.47

rency

-0.47

 Sahara

-0.47

hiro

-0.47

POSITIVE LOGITS

ASE

0.61

idge

0.59

ãĤ¡

0.58

vell

0.58

vice

0.58

vv

0.57

info

0.56

apest

0.54

à¤¾

0.53

es

0.53

Activations Density 0.250%

words related to challenges or problematic situations

words related to conflict and contention

No Comments

No Known Activations