INDEX

Explanations

distinct technical terms and identifiers from various domains, including programming, biology, and automotive references

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

Field groups

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

expandindo

-0.68

 propOrder

-0.59

twimg

-0.59

 referrerpolicy

-0.57

 AssemblyCulture

-0.57

+#+#

-0.54

principalColumn

-0.53

 المعيارى

-0.53

ftagPool

-0.53

 ujednoznacz

-0.53

POSITIVE LOGITS

 verst

0.44

 Deposit

0.42

istus

0.40

ftar

0.40

orem

0.39

 stru

0.39

 deposit

0.38

}{*}{

0.38

👈

0.37

Deposit

0.37

Activations Density 0.054%

distinct technical terms and identifiers from various domains, including programming, biology, and automotive references

Field groups

No Comments

No Known Activations