INDEX

Explanations

words related to physical actions or activities

oai_token-act-pair · gpt-3.5-turbo

phrases related to speed or performance metrics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 LGBTQ

-0.64

 Sharia

-0.64

 LGBT

-0.60

 hijab

-0.60

 transgender

-0.57

 Pepe

-0.53

 lesbian

-0.51

LGBT

-0.51

hammad

-0.51

 Koran

-0.51

POSITIVE LOGITS

lier

0.65

(>

0.63

pse

0.63

Bench

0.62

 Interstellar

0.60

EStream

0.60

 outl

0.60

pload

0.59

ivably

0.58

GHC

0.58

Activations Density 1.941%

words related to physical actions or activities

phrases related to speed or performance metrics

No Comments

No Known Activations