INDEX

Explanations

locations or landmarks

oai_token-act-pair · gpt-3.5-turbo

names of people, places, and specific 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

theless

-0.67

æĢ

-0.63

åĳ

-0.59

éĹ

-0.58

 reckoned

-0.56

CRIP

-0.55

ODUCT

-0.55

ulatory

-0.54

èª

-0.53

PDATE

-0.53

POSITIVE LOGITS

eworks

0.71

itars

0.70

antam

0.64

chlor

0.62

tones

0.61

ymes

0.59

ata

0.57

bane

0.57

utsch

0.56

rower

0.56

Activations Density 0.547%

locations or landmarks

names of people, places, and specific entities

No Comments

No Known Activations