INDEX

Explanations

mentions of specific geographical areas or regions

oai_token-act-pair · gpt-3.5-turbo

references to 'area' in various contexts, particularly in relation to discussions of safety or concern

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

inn

-0.77

arus

-0.75

estamp

-0.73

aldi

-0.73

bones

-0.73

å£«

-0.72

ANCE

-0.72

enstein

-0.71

tails

-0.69

ilon

-0.67

POSITIVE LOGITS

 surrounding

0.94

 bounded

0.91

 frequ

0.85

 encomp

0.83

 inhabited

0.78

 around

0.76

 nearest

0.76

rug

0.76

 adjacent

0.75

 where

0.74

Activations Density 0.062%

mentions of specific geographical areas or regions

references to 'area' in various contexts, particularly in relation to discussions of safety or concern

No Comments

No Known Activations