INDEX

Explanations

instances where the speaker refers to themselves or expresses personal opinions

oai_token-act-pair · gpt-3.5-turbo

references to personal experiences or statements

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

 Rolls

-0.64

 selves

-0.62

 Mats

-0.61

 careers

-0.61

 Georgian

-0.61

Kag

-0.58

 Sith

-0.58

edient

-0.58

 noses

-0.57

cedes

-0.56

POSITIVE LOGITS

'm

1.51

am

1.26

've

1.25

 myself

1.02

 adore

0.97

verson

0.96

RL

0.93

aido

0.93

 wish

0.87

'd

0.86

Activations Density 0.103%

instances where the speaker refers to themselves or expresses personal opinions

references to personal experiences or statements

No Comments

No Known Activations