INDEX

Explanations

occurrences of references to specific items or procedures

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

sentence start

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

 kasarigan

-0.45

󠁣

-0.43

documentElement

-0.39

UIControlState

-0.39

ventud

-0.39

ForValue

-0.38

 orgullo

-0.38

Prism

-0.38

fillType

-0.37

prism

-0.37

POSITIVE LOGITS

 initComponents

0.64

IntoConstraints

0.57

IContainer

0.47

CodedInputStream

0.46

awtextra

0.44

contentLoaded

0.44

 createSlice

0.42

 propOrder

0.42

tonode

0.41

 Followed

0.41

Activations Density 0.109%

occurrences of references to specific items or procedures

sentence start

No Comments

No Known Activations