INDEX

Explanations

phrases indicating a comparison or evaluation

oai_token-act-pair · gpt-3.5-turbo

phrases indicating speculation or conjecture

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

beh

-0.83

inav

-0.81

 teasp

-0.77

ãĤ§

-0.76

iHUD

-0.76

reason

-0.71

arse

-0.71

KNOWN

-0.70

 behavi

-0.70

iken

-0.69

POSITIVE LOGITS

 President

0.78

 celebrities

0.76

 Americans

0.75

 Narendra

0.74

 embattled

0.73

 Britons

0.72

 Britain

0.71

 politicians

0.71

 automakers

0.70

 billionaire

0.70

Activations Density 0.614%

phrases indicating a comparison or evaluation

phrases indicating speculation or conjecture

No Comments

No Known Activations