INDEX

Explanations

phrases indicating positive evaluation of someone's performance

oai_token-act-pair · gpt-3.5-turbo

phrases indicating the quality of performance 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 @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

Tang

-0.82

NetMessage

-0.77

fty

-0.65

eteria

-0.62

eros

-0.62

Phi

-0.62

isks

-0.62

ities

-0.61

ater

-0.60

raltar

-0.60

POSITIVE LOGITS

manship

0.94

 portraying

0.90

 ethic

0.83

 protecting

0.82

 illustrating

0.81

 educating

0.81

 showcasing

0.80

 combating

0.80

 handling

0.78

 translating

0.78

Activations Density 0.043%

phrases indicating positive evaluation of someone's performance

phrases indicating the quality of performance in various contexts

No Comments

No Known Activations