INDEX

Explanations

names, specifically surnames

oai_token-act-pair · gpt-3.5-turbo

the word "aut" and variations of it in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

BLE

-0.65

WAR

-0.59

soever

-0.58

birth

-0.56

WAY

-0.56

MIT

-0.55

TeX

-0.54

Uk

-0.54

function

-0.54

driving

-0.53

POSITIVE LOGITS

ilus

1.11

ista

0.89

ica

0.88

oir

0.87

ious

0.86

reau

0.85

urities

0.85

ilan

0.84

icit

0.81

erie

0.80

Activations Density 0.038%

names, specifically surnames

the word "aut" and variations of it in various contexts

No Comments

No Known Activations