INDEX

Explanations

mentions of police or military ranks

oai_token-act-pair · gpt-3.5-turbo

mentions of military ranks, specifically "Sgt" (Sergeant)

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ahime

-0.80

theless

-0.78

rights

-0.75

å¥³

-0.73

phal

-0.71

stage

-0.70

FORE

-0.68

hur

-0.66

topic

-0.65

POSITIVE LOGITS

geant

1.24

Sgt

1.13

 Sergeant

0.92

gt

0.92

 sergeant

0.90

veland

0.83

 Pepper

0.82

Maj

0.79

imen

0.79

illery

0.78

Activations Density 0.020%

mentions of police or military ranks

mentions of military ranks, specifically "Sgt" (Sergeant)

No Comments

No Known Activations