INDEX

Explanations

backslashes in various contexts, indicating file paths, escape sequences, or string formatting

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

dtrace probes

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

ôtel

-0.36

 Infórmanos

-0.34

rap

-0.34

SAL

-0.34

 obraz

-0.33

imshow

-0.33

 createState

-0.32

 Accredited

-0.32

GRect

-0.32

 kaynağından

-0.32

POSITIVE LOGITS

httphttps

0.62

SequentialGroup

0.62

__':

0.59

jsxFileName

0.59

__":

0.57

]")]

0.56

MLLoader

0.54

setVerticalGroup

0.51

 AssemblyCulture

0.51

AccessorTable

0.51

Activations Density 0.126%

backslashes in various contexts, indicating file paths, escape sequences, or string formatting

dtrace probes

No Comments

No Known Activations