INDEX

Explanations

phrases with contractions involving "won't"

oai_token-act-pair · gpt-3.5-turbo

phrases suggesting disbelief or denial

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Palest

-0.83

RAD

-0.80

 mosaic

-0.74

 Gleaming

-0.73

ctors

-0.73

pmwiki

-0.72

anwhile

-0.70

 horizont

-0.64

Buyable

-0.63

{:

-0.62

POSITIVE LOGITS

Â¢

0.97

¬

0.95

£

0.94

¼

0.91

½

0.91

¹

0.90

¿

0.90

ı

0.89

ł

0.87

»

0.86

Activations Density 0.167%

phrases with contractions involving "won't"

phrases suggesting disbelief or denial

No Comments

No Known Activations