INDEX

Explanations

the word "robot" or related terms

oai_token-act-pair · gpt-3.5-turbo

keywords related to robots and robotics

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

clair

-0.79

WAYS

-0.78

creen

-0.75

dn

-0.72

emi

-0.71

Beg

-0.70

tg

-0.70

arah

-0.69

lv

-0.68

tin

-0.68

POSITIVE LOGITS

ically

0.85

anical

0.82

 swarm

0.76

 crawling

0.75

 robot

0.72

 locom

0.72

arm

0.71

 overl

0.70

ocalypse

0.70

 bots

0.69

Activations Density 0.035%

the word "robot" or related terms

keywords related to robots and robotics

No Comments

No Known Activations