INDEX

Explanations

phrases related to covering up something or hiding information

oai_token-act-pair · gpt-3.5-turbo

phrases related to cover-ups and concealment

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 recomm

-0.77

Äĩ

-0.72

ovie

-0.69

 acquaintance

-0.68

uchi

-0.65

 sugg

-0.64

enos

-0.63

 Archdemon

-0.63

uckland

-0.63

ampa

-0.62

POSITIVE LOGITS

eatures

0.64

lain

0.62

 gaping

0.61

dden

0.60

 Qian

0.60

 bases

0.58

checks

0.58

tnc

0.57

topic

0.57

 flap

0.57

Activations Density 0.134%

phrases related to covering up something or hiding information

phrases related to cover-ups and concealment

No Comments

No Known Activations