INDEX

Explanations

mentions of combat-related words and phrases

oai_token-act-pair · gpt-3.5-turbo

references to military combat

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

Negative Logits

gow

-0.78

ħĭ

-0.73

eus

-0.69

bye

-0.68

Else

-0.65

ocument

-0.65

ingen

-0.65

 Seym

-0.64

ocl

-0.62

Choice

-0.61

POSITIVE LOGITS

iveness

0.99

ants

0.88

fighting

0.87

ant

0.86

 fatig

0.85

ments

0.84

ting

0.82

halla

0.76

 prowess

0.75

fighters

0.75

Activations Density 0.026%

mentions of combat-related words and phrases

references to military combat

No Known Activations