INDEX

Explanations

mentions of political figures or politically charged language

oai_token-act-pair · gpt-3.5-turbo

references to political terminology or discussions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 Warrant

-0.73

 Carbuncle

-0.69

 pity

-0.67

 Moor

-0.65

Ved

-0.64

MQ

-0.63

 fluorescent

-0.62

 recall

-0.62

URN

-0.61

 warrants

-0.61

POSITIVE LOGITS

icians

1.53

ician

1.45

ifact

1.27

ically

1.18

icial

1.14

eness

1.07

ileaks

0.99

ique

0.96

icking

0.91

cn

0.90

Activations Density 0.031%

mentions of political figures or politically charged language

references to political terminology or discussions

No Comments

No Known Activations