INDEX

Explanations

percentages and statistics

oai_token-act-pair · gpt-3.5-turbo

quantitative data and statistics

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

DonaldTrump

-0.48

ibrary

-0.48

inse

-0.44

gerald

-0.43

yip

-0.43

luster

-0.43

 Bulgar

-0.42

oulos

-0.41

 TheNitromeFan

-0.41

herer

-0.41

POSITIVE LOGITS

 thereof

0.85

 alike

0.80

 accordingly

0.79

 thereafter

0.73

 thereto

0.70

 therein

0.69

 respectively

0.56

 versa

0.56

 attRot

0.49

cellaneous

0.44

Activations Density 2.360%

percentages and statistics

quantitative data and statistics

No Comments

No Known Activations