INDEX

Explanations

references to a specific term or concept, probably "Horn" or related terms

oai_token-act-pair · gpt-3.5-turbo

references to the word "horn" and its variations in context

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

NCT

-0.90

arnaev

-0.86

PsyNetMessage

-0.84

ASED

-0.73

FORM

-0.68

ENCY

-0.68

ACT

-0.68

displayText

-0.66

-0.63

edia

-0.62

POSITIVE LOGITS

 Horn

1.34

horn

1.01

indal

0.94

pipe

0.93

beck

0.91

atility

0.88

qv

0.88

sey

0.84

atile

0.84

worms

0.82

Activations Density 0.009%

references to a specific term or concept, probably "Horn" or related terms

references to the word "horn" and its variations in context

No Comments

No Known Activations