INDEX

Explanations

keywords related to academic disciplines, awards, and research positions

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

academic fields and awards

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 nakalista

-0.63

 للاسماء

-0.63

fjspx

-0.60

ftagPool

-0.56

 leaſt

-0.54

 kaarangay

-0.54

 ſte

-0.54

 tranſ

-0.52

AddTagHelper

-0.52

 Мексичка

-0.52

POSITIVE LOGITS

by

0.39

in

0.36

 from

0.35

at

0.33

 members

0.32

 наоборот

0.32

 elsewhere

0.31

Примечание

0.31

ロリ

0.29

 else

0.29

Activations Density 0.050%

keywords related to academic disciplines, awards, and research positions

academic fields and awards

No Comments

No Known Activations