INDEX

Explanations

code-related terms associated with constants and variable declarations

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

menu, strip, offsets

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

MessageTagHelper

-0.59

ResumeLayout

-0.44

 protoimpl

-0.42

ApiModelProperty

-0.39

ViewFeatures

-0.38

izr

-0.38

rrggbb

-0.38

ExecuteReader

-0.38

 tän

-0.37

AccessorTable

-0.37

POSITIVE LOGITS

TryGetValue

0.45

 betweenstory

0.40

 undergoing

0.40

 Quiz

0.39

Gün

0.39

 namin

0.39

новништво

0.38

 encom

0.37

 مواليد

0.37

Quiz

0.36

Activations Density 0.036%

code-related terms associated with constants and variable declarations

menu, strip, offsets

No Comments

No Known Activations