INDEX

Explanations

phrases indicating importance or significance

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate singular identification or significance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ooks

-0.66

 cores

-0.64

ãĤµ

-0.62

 Palestin

-0.60

 fences

-0.59

inas

-0.58

older

-0.58

 warranties

-0.58

 cliffs

-0.58

inity

-0.58

POSITIVE LOGITS

 Hundred

0.91

 hundred

0.89

 dimensional

0.86

rency

0.81

 Thousand

0.75

 thing

0.74

 Piece

0.73

anchester

0.73

eree

0.72

 thousand

0.72

Activations Density 0.149%

phrases indicating importance or significance

phrases that indicate singular identification or significance

No Comments

No Known Activations