INDEX

Explanations

references to research findings or results in scientific texts

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

findings and results

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

 wireType

-0.73

 فريبيس

-0.70

 houſe

-0.69

 maș

-0.66

delwed

-0.65

ſelves

-0.64

 myſelf

-0.63

ſelf

-0.62

 verſ

-0.62

ⓧ

-0.62

POSITIVE LOGITS

 findings

1.02

 Findings

0.96

Findings

0.94

 FINDINGS

0.70

 results

0.66

 result

0.63

 hallaz

0.62

result

0.58

 Ergebnisse

0.57

find

0.57

Activations Density 0.024%

references to research findings or results in scientific texts

findings and results

No Comments

No Known Activations