INDEX

Explanations

requests and expressions of desire for specific outputs or actions

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

example, shown, eg

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

 CURIAM

-0.39

RegressionTest

-0.36

HttpEntity

-0.35

 élector

-0.34

viewBox

-0.32

BURGH

-0.32

CardBody

-0.31

सन्दर्भ

-0.30

viewDidLoad

-0.29

 FishBase

-0.28

POSITIVE LOGITS

WebElementEntity

0.56

Tracce

0.56

NewGuid

0.54

 Exacts

0.53

ulongan

0.52

 Photocase

0.51

hyrchwyd

0.51

MessageOf

0.49

ésultat

0.48

Ideally

0.48

Activations Density 0.130%

requests and expressions of desire for specific outputs or actions

example, shown, eg

No Comments

No Known Activations