INDEX

Explanations

statements or sentences that indicate user input or interaction

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

technical explanation or foreign language

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

AccessorTable

-0.66

GraphicsUnit

-0.63

 محفوظة

-0.62

 utafitiHapana

-0.61

хівовано

-0.60

tagHelperRunner

-0.60

 MainAxisSize

-0.60

DoubleQuotes

-0.60

+#+#

-0.58

RTCK

-0.58

POSITIVE LOGITS

 réparer

0.35

 circula

0.35

 enfans

0.34

 cebo

0.34

 perpé

0.34

 geç

0.33

 numériques

0.33

 économiques

0.33

 sürü

0.32

 voud

0.32

Activations Density 0.041%

statements or sentences that indicate user input or interaction

technical explanation or foreign language

No Comments

No Known Activations