INDEX

Explanations

references to political power structures or entities

oai_token-act-pair · gpt-3.5-turbo

variations of the word "put."

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

#$

-0.72

 compr

-0.72

alty

-0.64

issance

-0.63

 continents

-0.62

peak

-0.60

 purified

-0.60

 ende

-0.59

Ħ¢

-0.59

 Sins

-0.58

POSITIVE LOGITS

ierrez

1.41

tle

1.19

anamo

1.07

iful

1.04

ifully

0.99

ted

0.95

opian

0.94

opia

0.92

ting

0.90

cheon

0.88

Activations Density 0.011%

references to political power structures or entities

variations of the word "put."

No Comments

No Known Activations