INDEX

Explanations

phrases related to political or governmental actions

oai_token-act-pair · gpt-3.5-turbo

references to a specific symbol or character in various contexts

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

 Bris

-0.71

 Sapphire

-0.70

 confines

-0.66

 Bengal

-0.66

jri

-0.66

 shroud

-0.64

 guiActiveUnfocused

-0.64

 detached

-0.63

fid

-0.63

 indo

-0.62

POSITIVE LOGITS

¹

1.13

Ĳ

1.11

ª

1.08

ı

1.06

Ĵ

1.06

ł

1.05

ĳ

1.03

£

0.96

³

0.92

Į

0.90

Activations Density 0.133%

phrases related to political or governmental actions

references to a specific symbol or character in various contexts

No Comments

No Known Activations