INDEX

Explanations

phrases expressing extreme opinions or experiences

oai_token-act-pair · gpt-3.5-turbo

phrases emphasizing personal experiences and evaluations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

phasis

-0.60

gravity

-0.59

uish

-0.54

 dominates

-0.53

inks

-0.53

 Majority

-0.51

flix

-0.51

 Kung

-0.51

tiny

-0.51

papers

-0.50

POSITIVE LOGITS

 ever

1.57

 EVER

1.38

ever

1.13

 imaginable

1.07

 Ever

1.02

Ever

0.93

 encountered

0.92

 encount

0.85

 possibly

0.77

 anywhere

0.76

Activations Density 0.105%

phrases expressing extreme opinions or experiences

phrases emphasizing personal experiences and evaluations

No Comments

No Known Activations