INDEX

Explanations

phrases or commands related to adding items or information

oai_token-act-pair · gpt-3.5-turbo

instances of the word "Add" in relation to various actions or items

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

creen

-0.70

Ãį

-0.70

 writ

-0.65

 Gibbs

-0.64

Gel

-0.64

 Pulse

-0.64

DEM

-0.62

NING

-0.61

Bor

-0.61

ãĥīãĥ©ãĤ´ãĥ³

-0.61

POSITIVE LOGITS

endum

1.30

ition

1.26

itions

1.17

resses

1.16

icted

1.07

ictive

1.02

itionally

1.02

icts

1.00

itional

1.00

yson

0.99

Activations Density 0.018%

phrases or commands related to adding items or information

instances of the word "Add" in relation to various actions or items

No Comments

No Known Activations