INDEX

Explanations

proper nouns, especially related to a particular person named Ellison

oai_token-act-pair · gpt-3.5-turbo

mentions of the name "Ellison."

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

 Smoking

-0.71

ework

-0.70

Ľ

-0.69

 Vlad

-0.65

 Depend

-0.64

 Taiwan

-0.63

 START

-0.63

fully

-0.62

 Substance

-0.61

rolog

-0.60

POSITIVE LOGITS

 Ellison

1.15

irez

0.81

entanyl

0.72

urations

0.70

 Jinping

0.70

eering

0.69

 Tanz

0.68

worms

0.68

 mosquit

0.68

 lapt

0.67

Activations Density 0.005%

proper nouns, especially related to a particular person named Ellison

mentions of the name "Ellison."

No Comments

No Known Activations