INDEX

Explanations

phrases indicating personal preferences or experiences

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

actions 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

StoryboardSegue

-0.48

WireFormatLite

-0.48

WriteLiteral

-0.47

 ciudadana

-0.47

tagHelperRunner

-0.47

//

-0.43

 للاسماء

-0.43

theless

-0.42

LEGGI

-0.42

 pubblici

-0.42

POSITIVE LOGITS

mis

0.47

httphttps

0.44

 boss

0.44

key

0.43

Met

0.43

word

0.42

rette

0.41

qtype

0.41

ilent

0.40

Gle

0.40

Activations Density 0.009%

phrases indicating personal preferences or experiences

actions followed by actions

No Comments

No Known Activations