INDEX

Explanations

proper nouns or specific named entities

oai_token-act-pair · gpt-3.5-turbo

references to specific geographic locations or terms related to geography

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

oku

-0.66

hands

-0.60

 Azerbaijan

-0.60

onym

-0.60

avoid

-0.59

sylv

-0.59

fig

-0.58

tsky

-0.58

guide

-0.58

 stub

-0.58

POSITIVE LOGITS

enhagen

0.82

owship

0.77

andise

0.76

ioxide

0.74

ruary

0.70

imately

0.69

elli

0.68

iously

0.67

coni

0.66

ulin

0.66

Activations Density 0.420%

proper nouns or specific named entities

references to specific geographic locations or terms related to geography

No Comments

No Known Activations