INDEX

Explanations

words related to newness or novelty

oai_token-act-pair · gpt-3.5-turbo

references to new beginnings or new entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

mination

-0.87

ashtra

-0.79

jar

-0.79

halla

-0.76

ostic

-0.74

xual

-0.74

Zip

-0.73

chnology

-0.72

rary

-0.72

 Hayward

-0.71

POSITIVE LOGITS

bies

0.91

bie

0.89

 arrivals

0.87

 teammates

0.85

 lease

0.84

 batch

0.83

 surroundings

0.82

 acquaintances

0.82

 venture

0.82

 predicament

0.81

Activations Density 0.046%

words related to newness or novelty

references to new beginnings or new entities

No Comments

No Known Activations