INDEX

Explanations

proper nouns, specifically names of individuals

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names of individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

illy

-0.81

agra

-0.78

apeshifter

-0.77

information

-0.71

abol

-0.70

avers

-0.70

file

-0.69

creen

-0.69

eeper

-0.69

elling

-0.68

POSITIVE LOGITS

sson

1.48

 Wilhelm

1.20

 Hitler

1.20

 Schwarzenegger

1.16

 Herz

1.12

 Engels

1.11

von

1.08

Sch

1.07

 Ludwig

1.07

 Schwarz

1.06

Activations Density 0.148%

proper nouns, specifically names of individuals

proper nouns, particularly names of individuals

No Comments

No Known Activations