INDEX

Explanations

phrases that indicate purpose or intent related to actions

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

Comes after the word "for"

np_acts-logits-general · gemini-2.0-flash

allows for

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

Dispose

-0.40

 jurisdic

-0.40

 people

-0.40

 because

-0.38

SetPosition

-0.37

Comer

-0.37

 Helen

-0.37

 Bere

-0.36

ReadOnly

-0.36

 philosop

-0.35

POSITIVE LOGITS

 الحره

0.81

 utafitiHapana

0.75

migrationBuilder

0.73

extAlignment

0.73

SBATCH

0.72

 surla

0.71



0.70

ſicht

0.70

 NSCoder

0.69

principalColumn

0.69

Activations Density 0.432%

phrases that indicate purpose or intent related to actions

Comes after the word "for"

allows for

No Comments

No Known Activations