INDEX

Explanations

questions and inquiries related to understanding or identifying information

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

discussing reviews or foreign concepts

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

ecap

-0.42

DoubleQuotes

-0.42

#+#

-0.41

rungsseite

-0.40

gies

-0.40

/*---

-0.38

rtype

-0.38

ssz

-0.37

 otomatig

-0.37



-0.35

POSITIVE LOGITS

 समीक्षक

0.49

 almend

0.49

 prohibido

0.45

MessageOf

0.44

 itse

0.44

 âme

0.43

themselves

0.43

 itself

0.43

itself

0.42

 themselves

0.42

Activations Density 0.128%

questions and inquiries related to understanding or identifying information

discussing reviews or foreign concepts

No Comments

No Known Activations