INDEX

Explanations

specific terms related to electrical and biochemical processes in the context of neurons

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

binding activity Recovery gain prejudice

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

-0.46

htm

-0.41

 Garner

-0.39

 Hatch

-0.38

itoral

-0.37

 Bondi

-0.37

IsPostBack

-0.37

hat

-0.37

-0.36

 vergeten

-0.36

POSITIVE LOGITS

 صوتيه

0.61

 disambiguazione

0.61

aarrggbb

0.60

MLLoader

0.57

Diweddarwch

0.55

ⓧ

0.53

 препратки

0.52

 فريبيس

0.51

:✨

0.50

 informée

0.50

Activations Density 0.067%

specific terms related to electrical and biochemical processes in the context of neurons

binding activity Recovery gain prejudice

No Comments

No Known Activations