INDEX

Explanations

phrases indicating a conclusion or judgment

oai_token-act-pair · gpt-3.5-turbo

phrases indicating safety or certainty

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ious

-0.61

unity

-0.61

 Ended

-0.60

viol

-0.60

ucha

-0.60

forms

-0.59

*/(

-0.59

tons

-0.58

naires

-0.58

rawdownloadcloneembedreportprint

-0.58

POSITIVE LOGITS

say

1.94

SAY

1.52

 saying

1.33

say

1.31

 conclude

1.27

 admit

1.22

 remark

1.21

 concede

1.20

 suggest

1.19

 declare

1.12

Activations Density 0.191%

phrases indicating a conclusion or judgment

phrases indicating safety or certainty

No Comments

No Known Activations