INDEX

Explanations

references to law enforcement officers or agencies with the word "Special" followed by a high-ranking descriptor

oai_token-act-pair · gpt-3.5-turbo

references to special operations and military entities

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

•Special

No Comments

Negative Logits

iHUD

-0.69

 Twain

-0.69

kj

-0.68

-0.66

GGGG

-0.62

ãĤ¼

-0.61

ILLE

-0.61

shown

-0.60

\\\\\\\\

-0.60

vu

-0.59

POSITIVE LOGITS

izations

1.04

ised

0.98

ized

0.98

isations

0.91

Ops

0.90

ties

0.89

ization

0.87

 Interest

0.87

izes

0.86

ities

0.84

Activations Density 0.022%

references to law enforcement officers or agencies with the word "Special" followed by a high-ranking descriptor

references to special operations and military entities

No Comments

No Known Activations