INDEX

Explanations

semicolons and expressions involving results in code

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

code result variable

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

openqa

-0.49

IsContent

-0.43

NameInMap

-0.42

DrawerToggle

-0.40

UniformLocation

-0.35

FromArgb

-0.35

 AssemblyCulture

-0.34

 results

-0.34

matchCondition

-0.34

sendRedirect

-0.34

POSITIVE LOGITS

tisgarh

0.53

enderror

0.50

 resourceCulture

0.48

羔

0.47

 IFTT

0.47

 règlement

0.46

exitRule

0.45

 Messina

0.45

posób

0.44

iddhar

0.44

Activations Density 0.012%

semicolons and expressions involving results in code

code result variable

No Comments

No Known Activations