INDEX

Explanations

programming constructs and statuses associated with error handling and component management

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

attributes, types, and resources

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

 AssemblyCompany

-0.42

inSlope

-0.37

working

-0.37

########.

-0.36

 safely

-0.35

従

-0.34

PYX

-0.34

끕

-0.34

 referenties

-0.33

AddTagHelper

-0.33

POSITIVE LOGITS

InjectAttribute

0.53

boutin

0.52

 NDEBUG

0.47

 HttpNotFound

0.46

ophylla

0.46

 melainkan

0.45

 אלא

0.45

jsxFileName

0.43

ValueGeneration

0.41

 للمعارف

0.40

Activations Density 0.059%

programming constructs and statuses associated with error handling and component management

attributes, types, and resources

No Comments

No Known Activations