INDEX

Explanations

repeated patterns or sequences in strings

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

quoted strings and concatenation

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

 bookId

-0.72

addGap

-0.59

bindingNavigator

-0.58

 ganglia

-0.57

[toxicity=0]

-0.57

 noqa

-0.57

 carb

-0.57

thenia

-0.56

 CreateTagHelper

-0.56

 Parton

-0.56

POSITIVE LOGITS

 didalam

0.63

 diatas

0.59

(""+

0.54

/"+

0.52

insegna

0.50

 diantaranya

0.48

 fjer

0.48

 dibawah

0.47

ofthe

0.47

":"+

0.47

Activations Density 0.003%

repeated patterns or sequences in strings

quoted strings and concatenation

No Comments

No Known Activations