INDEX

Explanations

affirmative responses

oai_token-act-pair · gpt-3.5-turbo

affirmative responses or confirmations

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

bage

-0.82

RAW

-0.80

perial

-0.77

actionDate

-0.70

 ILCS

-0.69

lines

-0.69

 externalToEVAOnly

-0.67

leted

-0.66

posted

-0.66

liner

-0.65

POSITIVE LOGITS

terday

1.75

sir

0.91

yes

0.80

YES

0.78

ZI

0.77

YES

0.77

yes

0.72

Yes

0.71

yne

0.71

 Pryor

0.70

Activations Density 0.020%

affirmative responses

affirmative responses or confirmations

No Comments

No Known Activations