INDEX

Explanations

phrases containing the word 'con' or 'uncon'

oai_token-act-pair · gpt-3.5-turbo

terms related to conformity and unconventionality

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Twice

-0.65

assetsadobe

-0.65

...]

-0.65

 Rover

-0.63

BILITIES

-0.63

 Rated

-0.63

chens

-0.62

 scratch

-0.61

TPS

-0.58

BILITY

-0.58

POSITIVE LOGITS

ventions

1.11

stant

1.07

con

1.01

currency

0.99

vict

0.95

crete

0.94

vention

0.92

secut

0.91

clus

0.90

rad

0.89

Activations Density 0.006%

phrases containing the word 'con' or 'uncon'

terms related to conformity and unconventionality

No Comments

No Known Activations