INDEX

Explanations

sequences of repeated characters or whitespace patterns

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

template syntax and types

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

Autoritní

-0.65

 morire

-0.47

PostMapping

-0.46

 bezeichneter

-0.46

RuleContext

-0.45

حوالہ

-0.44

 fusca

-0.44

Kaynakça

-0.44

 Prozesse

-0.43

 semilla

-0.43

POSITIVE LOGITS

 Lana

0.43

fortawesome

0.43

__*/

0.42

phem

0.42

 concrete

0.41

 FetchType

0.41

よって

0.41

 watercolor

0.40

usercontent

0.39

Activations Density 0.037%

sequences of repeated characters or whitespace patterns

template syntax and types

No Comments

No Known Activations