INDEX

Explanations

specific numerical data or identifiers in the text

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

citation or image references

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

 kiệm

-0.50

 financière

-0.44

клопе

-0.42

 telefónica

-0.42

 nueces

-0.41

 Signalez

-0.41

ttemberg

-0.40

醐

-0.40

DIPSETTING

-0.40

 económica

-0.40

POSITIVE LOGITS

fjspx

0.52

DoubleQuotes

0.39

Same

0.38

UniformLocation

0.37

NSIndexPath

0.37

ViewFeatures

0.37

 AssemblyCompany

0.36

AddWithValue

0.36

 CanadaChoose

0.36

amoan

0.35

Activations Density 0.013%

specific numerical data or identifiers in the text

citation or image references

No Comments

No Known Activations