INDEX

Explanations

specific titles and names associated with various categories or types, including books, films, and other media

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

features, rules, or stages

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

 właśnie

-0.35

 Diakses

-0.33

Kanpo

-0.33

Pozdrawiam

-0.33

↵

-0.33

<strong>

-0.30

ścioła

-0.30

 entera

-0.29

likle

-0.29

 oglą

-0.29

POSITIVE LOGITS

ſelben

0.75

ſammen

0.68

󠀠

0.66

 nonUne

0.66

vorschaubild

0.64

wiliwch

0.63

⪜

0.63

 zuſammen

0.61

WriteTagHelper

0.60

 queſto

0.59

Activations Density 0.184%

specific titles and names associated with various categories or types, including books, films, and other media

features, rules, or stages

No Comments

No Known Activations