INDEX

Explanations

people being interviewed or having conversations

oai_token-act-pair · gpt-3.5-turbo

phrases indicating conversations or discussions with individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

interstitial

-0.80

inventoryQuantity

-0.75

 inflamm

-0.64

)</

-0.63

coming

-0.63

 hurting

-0.63

matter

-0.62

 endanger

-0.60

 collapsing

-0.59

ERAL

-0.59

POSITIVE LOGITS

inqu

0.77

eries

0.74

edom

0.73

 firsthand

0.72

odcast

0.72

 Chow

0.70

 extensively

0.70

 Eater

0.69

 Greenberg

0.69

 interview

0.68

Activations Density 0.304%

people being interviewed or having conversations

phrases indicating conversations or discussions with individuals

No Comments

No Known Activations