INDEX

Explanations

mentions of things being visually perceived or aesthetically described

oai_token-act-pair · gpt-3.5-turbo

references to visual appearances or aesthetics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 Literature

-0.74

cffff

-0.73

wine

-0.68

erto

-0.64

 Kurd

-0.64

 Tosh

-0.63

scribed

-0.62

acus

-0.61

icipated

-0.59

 Loch

-0.59

POSITIVE LOGITS

ahead

1.02

checks

0.79

finder

0.77

ENE

0.73

ups

0.73

tones

0.68

bones

0.66

etic

0.66

 looks

0.66

outs

0.66

Activations Density 0.037%

mentions of things being visually perceived or aesthetically described

references to visual appearances or aesthetics

No Comments

No Known Activations