INDEX

Explanations

references to the location "Kawagawa."

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "awa" across various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

tone

-0.71

book

-0.69

 troopers

-0.65

Ð¾Ð

-0.64

urgy

-0.63

 OECD

-0.63

pick

-0.62

hematically

-0.60

 Normandy

-0.60

test

-0.58

POSITIVE LOGITS

awa

1.14

aii

1.12

oka

0.83

orthy

0.81

ibur

0.81

imura

0.81

anan

0.80

 Shogun

0.80

qua

0.79

iba

0.79

Activations Density 0.016%

references to the location "Kawagawa."

occurrences of the term "awa" across various contexts

No Comments

No Known Activations