INDEX

Explanations

mentions of body parts and objects being held or placed

oai_token-act-pair · gpt-3.5-turbo

references to physical objects and their locations in relation to people

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

olutions

-0.73

 continents

-0.72

QL

-0.71

assadors

-0.68

 architects

-0.67

otech

-0.65

 Reloaded

-0.65

DEM

-0.65

Ã©t

-0.64

landish

-0.64

POSITIVE LOGITS

 pocket

1.34

 pockets

1.24

 vicinity

1.16

 midst

1.16

 hole

1.14

 holster

1.13

 drawer

1.10

 compartment

1.06

 cavity

1.06

 recess

1.04

Activations Density 0.197%

mentions of body parts and objects being held or placed

references to physical objects and their locations in relation to people

No Comments

No Known Activations