INDEX

Explanations

references to the self and personal thoughts, often starting with "I"

oai_token-act-pair · gpt-3.5-turbo

instances of the word "I," indicating a focus on personal expression and self-reference

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

rising

-0.63

marg

-0.61

arians

-0.59

 Uriel

-0.59

 Jarrett

-0.58

 Aberdeen

-0.56

 Pearson

-0.55

 Vald

-0.55

 Walton

-0.55

 Rubio

-0.54

POSITIVE LOGITS

'm

1.56

've

1.31

'll

1.13

RL

1.04

'd

1.02

verson

1.02

am

0.99

ggy

0.97

 suppose

0.96

KE

0.94

Activations Density 0.281%

references to the self and personal thoughts, often starting with "I"

instances of the word "I," indicating a focus on personal expression and self-reference

No Comments

No Known Activations