INDEX

Explanations

specific entities or identifiers, particularly related to unique items or examples

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

besides the other

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

/*---

-0.52

awaiter

-0.44

DockStyle

-0.42

 coverage

-0.42

meler

-0.41

classmethod

-0.41

 <>",

-0.40

 raso

-0.39

[]:

-0.39

WP

-0.39

POSITIVE LOGITS

parsedMessage

0.52

WillAppear

0.45

\{\\

0.45

SOUNDBITE

0.42

 CanadaChoose

0.41

न्दर्भ

0.41

woordig

0.40

RenderAtEndOf

0.40

RTLD

0.40

 sondern

0.40

Activations Density 0.035%

specific entities or identifiers, particularly related to unique items or examples

besides the other

No Comments

No Known Activations