INDEX

Explanations

phrases related to weapons and armed conflict

oai_token-act-pair · gpt-3.5-turbo

references to armaments or military equipment

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Vide

-0.72

 Sound

-0.67

 Dough

-0.66

 Atmosp

-0.64

 QuÃ©

-0.61

BET

-0.58

 elig

-0.58

 Nare

-0.57

ween

-0.56

 Chaff

-0.56

POSITIVE LOGITS

chair

1.53

aments

1.44

ament

1.44

ageddon

1.37

chairs

1.30

strong

1.07

oire

1.06

ando

1.05

orers

1.03

illary

1.02

Activations Density 0.023%

phrases related to weapons and armed conflict

references to armaments or military equipment

No Comments

No Known Activations