INDEX

Explanations

specific mentions of names or titles, likely in a formal information context

oai_token-act-pair · gpt-3.5-turbo

punctuation marks and abbreviations, particularly related to television networks or programs

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

hoff

-0.78

poke

-0.75

ahime

-0.71

 Feast

-0.70

owitz

-0.68

isite

-0.68

bery

-0.66

inger

-0.65

ensional

-0.64

zees

-0.64

POSITIVE LOGITS

ulhu

0.88

emporary

0.87

ottage

0.69

inelli

0.68

MX

0.68

 eleg

0.66

usting

0.66

UGC

0.64

ombat

0.64

Activations Density 0.068%

specific mentions of names or titles, likely in a formal information context

punctuation marks and abbreviations, particularly related to television networks or programs

No Comments

No Known Activations