INDEX

Explanations

punctuation marks, particularly periods and negative signs

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

numbers after 5

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

rayage

-0.47

crochet

-0.43

Routledge

-0.41

surat

-0.41

opengl

-0.41

 copertura

-0.40

DrawerLayout

-0.40

 mør

-0.40

 natten

-0.40

 rése

-0.39

POSITIVE LOGITS

1.06

 Fifth

0.82

 Five

0.82

 fifth

0.77

 five

0.77

Five

0.75

Fifth

0.73

 Cinco

0.73

 cinco

0.72

５

0.72

Activations Density 0.190%

punctuation marks, particularly periods and negative signs

numbers after 5

No Comments

No Known Activations