INDEX

Explanations

references to puzzles in various contexts

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

puzzles, maze, game

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

 berlebihan

-0.43

INCREF

-0.43

 vigueur

-0.42

ţilor

-0.41

 cenderung

-0.41

from

-0.41

 środków

-0.40

 ibunya

-0.40

ponses

-0.40

 wyłącznie

-0.40

POSITIVE LOGITS

 puzzle

1.27

 Puzzle

1.24

puzzle

1.19

Puzzle

1.14

 puzzles

1.06

 Puzzles

1.01

 puzz

0.87

ZZLE

0.78

JSONException

0.69

 ſind

0.69

Activations Density 0.006%

references to puzzles in various contexts

puzzles, maze, game

No Comments

No Known Activations