INDEX

Explanations

names of people or places

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names

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

bound

-0.73

bre

-0.67

breaking

-0.61

position

-0.60

ivated

-0.60

hillary

-0.57

stri

-0.57

breakers

-0.56

imm

-0.56

con

-0.56

POSITIVE LOGITS

's

0.70

 herself

0.69

Sr

0.67

 commented

0.64

stals

0.64

ites

0.63

iev

0.62

inen

0.61

enegger

0.61

KE

0.61

Activations Density 0.290%

names of people or places

proper nouns, particularly names

No Comments

No Known Activations