INDEX

Explanations

phrases and concepts related to intentions or purposes

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

intended to

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

Források

-0.47

 resourceCulture

-0.43

TRAILING

-0.39

 conf

-0.37

Einzelnachweise

-0.36

xer

-0.36

GALL

-0.36

conf

-0.36

 Sachen

-0.36

іод

-0.35

POSITIVE LOGITS

intended

1.05

 intended

1.02

Intended

0.96

 Intended

0.95

 planned

0.92

planned

0.90

 intending

0.81

 intention

0.80

 intends

0.80

 intend

0.79

Activations Density 0.097%

phrases and concepts related to intentions or purposes

intended to

No Comments

No Known Activations