INDEX

Explanations

numeric identifiers or references

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

numerical identifiers

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

 continuidad

-0.55

 confiar

-0.44

 esfuerzo

-0.43

 permiso

-0.42

 Finanzierung

-0.42

ỡng

-0.41

 operativos

-0.41

 abierta

-0.40

aksanaan

-0.40

 financiera

-0.40

POSITIVE LOGITS

 numéro

1.00

 nomor

0.96

 Nummer

0.95

 Number

0.94

 number

0.93

 номер

0.92

Номер

0.85

 NUMBER

0.85

Number

0.85

 numbers

0.82

Activations Density 0.012%

numeric identifiers or references

numerical identifiers

No Comments

No Known Activations