INDEX

Explanations

terms related to specific events or incidents being reported by eyewitnesses in news articles

oai_token-act-pair · gpt-3.5-turbo

references to legal and emergency services situations

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

().

-0.72

 âĢ¦"

-0.69

 ðŁĻĤ

-0.67

etc

-0.67

 ______

-0.64

âĢķ

-0.64

("

-0.64

.",

-0.64

âĻ¥

-0.63

Ïī

-0.63

POSITIVE LOGITS

ogether

1.00

icularly

0.95

inarily

0.94

izens

0.89

goers

0.88

 Details

0.83

quartered

0.81

ificantly

0.80

itionally

0.80

fighters

0.80

Activations Density 0.329%

terms related to specific events or incidents being reported by eyewitnesses in news articles

references to legal and emergency services situations

No Comments

No Known Activations