INDEX

Explanations

sections of code or markup that indicate structure or format

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

closing structural characters

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

 queſta

-1.22

iſchen

-0.99

Autoritní

-0.94

 vooz

-0.92

 Infórmanos

-0.91

ſammen

-0.91

parsedMessage

-0.90

linawan

-0.90

ſſung

-0.90

$_"

-0.90

POSITIVE LOGITS

1.36

1.16

0.98

0.93

0.91

0.89

0.88

0.87

0.85

Activations Density 0.037%

sections of code or markup that indicate structure or format

closing structural characters

No Comments

No Known Activations