INDEX

Explanations

mentions of products or projects labeled as "feature"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "feature."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

sis

-0.87

vernment

-0.76

apy

-0.75

uter

-0.74

si

-0.73

usable

-0.72

externalActionCode

-0.71

azar

-0.68

ears

-0.64

amn

-0.64

POSITIVE LOGITS

Feature

1.14

 features

1.07

 Features

1.07

feature

1.03

features

1.02

Features

1.01

 feature

0.95

 prominently

0.93

 Feature

0.84

ttes

0.84

Activations Density 0.010%

mentions of products or projects labeled as "feature"

occurrences of the word "feature."

No Comments

No Known Activations