INDEX

Explanations

symbols and syntax commonly found in programming or markup languages

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

object `->` `.`

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

 betweenstory

-0.77

principalColumn

-0.72

berdayakan

-0.66

 صوتيه

-0.65

principalTable

-0.63

 snippetHide

-0.62

tonode

-0.61

fromnode

-0.61

Personendaten

-0.60

Демографія

-0.59

POSITIVE LOGITS

fjspx

0.52

->

0.47

 Load

0.45

Loa

0.43

 Dade

0.42

 load

0.42

 Caribbean

0.41

 Jamb

0.41

Activations Density 0.000%

symbols and syntax commonly found in programming or markup languages

object `->` `.`

No Comments

No Known Activations