INDEX

Explanations

specific proper nouns, particularly names and locations associated with scientific or academic contexts

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

CETP or Tivoli Gardens

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

<bos>

-0.58

 deleteById

-0.57

-------

-0.52

vscode

-0.50

 Agassi

-0.48

Luigi

-0.47

↵↵↵

-0.46

IContainer

-0.44

<eos>

-0.44

POSITIVE LOGITS

BibitemShut

0.66

arangay

0.58

rsiniz

0.56

 fört

0.55

zionalità

0.55

rasında

0.54

CET

0.54

pulseira

0.54

 Sentry

0.51

attutto

0.51

Activations Density 0.200%

specific proper nouns, particularly names and locations associated with scientific or academic contexts

CETP or Tivoli Gardens

No Comments

No Known Activations