INDEX

Explanations

mentions of shops or shop-related activities

oai_token-act-pair · gpt-3.5-turbo

references to shops and retail environments

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

etus

-0.64

 Benedict

-0.61

 Humph

-0.58

 flakes

-0.58

POW

-0.58

 Couch

-0.58

 worse

-0.57

 trooper

-0.57

DragonMagazine

-0.57

ioxide

-0.56

POSITIVE LOGITS

lifting

1.44

keeper

1.44

keepers

1.42

worn

1.19

lift

1.12

front

1.11

keep

1.08

hops

1.03

shop

1.03

floor

1.03

Activations Density 0.065%

mentions of shops or shop-related activities

references to shops and retail environments

No Comments

No Known Activations