INDEX

Explanations

multiple instances of the word "assuming" and evaluate conditional statements

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

mathematical expressions and names

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

RectangleBorder

-0.57

InputBorder

-0.53

 direta

-0.50

 InputDecoration

-0.47



-0.44

astéroïdes

-0.44

 bluzka

-0.43

⟬

-0.43

contentLoaded

-0.42

 direto

-0.42

POSITIVE LOGITS

ModelAdmin

0.48

 otomatig

0.43

 tsunami

0.41

Remarks

0.41

Havolalar

0.40

CSO

0.40

 cools

0.39

 slowest

0.39

Bibliograf

0.38

JSONException

0.38

Activations Density 0.074%

multiple instances of the word "assuming" and evaluate conditional statements

mathematical expressions and names

No Comments

No Known Activations