INDEX

Explanations

comparisons or contrasts between different concepts or entities

oai_token-act-pair · gpt-3.5-turbo

phrases and words associated with negative attributes or conditions

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

doms

-0.64

sheets

-0.60

 administrations

-0.60

ipeg

-0.60

iots

-0.58

 coats

-0.58

 wills

-0.56

 backgrounds

-0.56

bats

-0.56

roots

-0.56

POSITIVE LOGITS

TBD

0.57

 enough

0.51

 Beir

0.51

omorphic

0.50

 Gibraltar

0.50

actly

0.50

omorph

0.50

 Cunningham

0.48

 synonymous

0.48

urated

0.48

Activations Density 1.324%

comparisons or contrasts between different concepts or entities

phrases and words associated with negative attributes or conditions

No Comments

No Known Activations