INDEX

Explanations

the word 'York' with varying degrees of activation

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "York."

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

cius

-0.79

eers

-0.75

anwhile

-0.73

itures

-0.70

aeda

-0.70

 Flavoring

-0.70

pmwiki

-0.68

phrine

-0.68

cedes

-0.67

igmat

-0.66

POSITIVE LOGITS

mares

0.88

York

0.87

dale

0.84

 Huss

0.77

stone

0.75

onne

0.73

cap

0.70

stones

0.70

 Brook

0.70

Activations Density 0.010%

the word 'York' with varying degrees of activation

occurrences of the word "York."

No Comments

No Known Activations