INDEX

Explanations

instances where someone expresses involvement or commitment to a particular situation

oai_token-act-pair · gpt-3.5-turbo

repetitions and references to the word "this."

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

ãĤ¹ãĥĪ

-0.92

idelines

-0.85

alties

-0.81

Ĭ±

-0.81

±

-0.79

Ĵ

-0.78

izontal

-0.78

sts

-0.78

asures

-0.78

·

-0.78

POSITIVE LOGITS

guy

1.29

 shit

1.15

 stuff

1.11

 thing

1.09

 crap

1.08

 sucker

1.05

 particular

1.03

 whole

1.01

 shitty

0.97

 entire

0.91

Activations Density 0.163%

instances where someone expresses involvement or commitment to a particular situation

repetitions and references to the word "this."

No Comments

No Known Activations