INDEX

Explanations

mentions of military combat and related terms

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 @ 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

gow

-0.80

Else

-0.73

eus

-0.68

ħĭ

-0.68

atron

-0.66

eton

-0.66

ocl

-0.65

ocument

-0.64

apply

-0.62

earch

-0.61

POSITIVE LOGITS

fighting

0.91

iveness

0.88

fighters

0.82

 fatig

0.81

halla

0.79

ants

0.78

 prowess

0.78

ting

0.78

ant

0.77

fighter

0.76

Activations Density 0.023%

mentions of military combat and related terms

references to military combat

No Comments

No Known Activations