INDEX

Explanations

cost-related metrics in legal or algorithmic contexts

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

numbers and comparisons

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

########.

-0.66

 مشين

-0.66

 kasarigan

-0.62

 mukana

-0.60

batore

-0.60

majánló

-0.60

 pinulongan

-0.59

RTLR

-0.58

UnknownFieldSet

-0.57

Ӕ

-0.56

POSITIVE LOGITS

 significantly

0.43

Rod

0.42

 regular

0.40

Rod

0.40

 normal

0.37

 versus

0.35

 anders

0.35

 otherwise

0.34

 compared

0.33

rod

0.33

Activations Density 0.576%

cost-related metrics in legal or algorithmic contexts

numbers and comparisons

No Comments

No Known Activations