INDEX

Explanations

words related to military or violent actions, specifically the word "bombing"

oai_token-act-pair · gpt-3.5-turbo

references to bombing events or campaigns

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

laus

-0.95

eva

-0.76

Lear

-0.75

learn

-0.75

mia

-0.74

BOOK

-0.74

ITY

-0.74

mis

-0.69

Tang

-0.69

dit

-0.68

POSITIVE LOGITS

 bombing

1.22

 bombings

1.06

 raids

1.03

 bomber

0.97

 spree

0.97

 barr

0.94

 bombers

0.91

 bombed

0.86

 raid

0.83

 bombard

0.82

Activations Density 0.017%

words related to military or violent actions, specifically the word "bombing"

references to bombing events or campaigns

No Comments

No Known Activations