INDEX

Explanations

references to visual elements and experiences

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

acts of seeing

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

 labourers

-0.52

 Menon

-0.51

loài

-0.51

ßig

-0.50

 Infórmanos

-0.50

 reformers

-0.48

Papers

-0.47

 Berner

-0.47

 Malo

-0.47

MPP

-0.47

POSITIVE LOGITS

 Sight

1.15

 sight

1.05

sight

0.98

Sight

0.96

 sights

0.78

 pleaſure

0.62

 sighting

0.61

sightly

0.60

 sightings

0.60

light

0.59

Activations Density 0.006%

references to visual elements and experiences

acts of seeing

No Comments

No Known Activations