INDEX

Explanations

occurrences of specific token patterns or formatting in data

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

dy notation in calculus

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

AndEndTag

-0.61

 Италијани

-0.57

RTEX

-0.57

 årene

-0.54

 zijne

-0.52

 szczeg

-0.52

AddTagHelper

-0.52

 Infór

-0.50

AnchorStyles

-0.50

 arbejde

-0.49

POSITIVE LOGITS

male

0.45

 random

0.45

 loud

0.44

random

0.44

 transcript

0.43

Random

0.43

 randomness

0.42

loud

0.42

 bruto

0.42

 Darm

0.42

Activations Density 0.000%

occurrences of specific token patterns or formatting in data

dy notation in calculus

No Comments

No Known Activations