INDEX

Explanations

instructions and procedural phrases related to tasks and actions

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

instructions followed by actions

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

-0.64

 ModelExpression

-0.62

Портали

-0.61

IsPostBack

-0.57

Portály

-0.57

Bibliograf

-0.56

IsUnicode

-0.54

 Administrativna

-0.54

ſelves

-0.52

webElementXpaths

-0.50

POSITIVE LOGITS

 simply

0.46

simply

0.43

DockStyle

0.40

RenderAtEndOf

0.36

別

0.35

by

0.35

 следу

0.31

either

0.31

 když

0.31

เพียง

0.31

Activations Density 0.060%

instructions and procedural phrases related to tasks and actions

instructions followed by actions

No Comments

No Known Activations