INDEX

Explanations

terms related to medical conditions and treatment efficacy

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

years of success

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

 learnt

-0.47

 ModelExpression

-0.45

etc

-0.40

 Persson

-0.37

ถม

-0.37

UnusedPrivate

-0.35

 firstly

-0.35

 Schäfer

-0.33

 acá

-0.33

ต่างๆ

-0.33

POSITIVE LOGITS

 Италијани

0.78

 ſtate

0.71

 fubject

0.69

 Савезне

0.69

 pleaſure

0.67

 kasarigan

0.66

 ſever

0.65

hyrchwyd

0.65

 houſe

0.65

 poffible

0.64

Activations Density 0.166%

terms related to medical conditions and treatment efficacy

years of success

No Comments

No Known Activations