INDEX

Explanations

proper nouns related to various topics, possibly indicating a focus on specific entities or concepts

oai_token-act-pair · gpt-3.5-turbo

proper nouns and specific entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

mel

-0.58

tains

-0.57

-0.56

$.

-0.56

 Sass

-0.55

 Stam

-0.55

(),

-0.55

irement

-0.55

:,

-0.54

 Loyal

-0.54

POSITIVE LOGITS

 alike

1.62

are

1.32

 were

1.26

 weren

1.18

 aren

1.18

 respectively

1.11

 have

1.00

 collide

0.99

 remain

0.96

were

0.94

Activations Density 0.556%

proper nouns related to various topics, possibly indicating a focus on specific entities or concepts

proper nouns and specific entities

No Comments

No Known Activations