INDEX

Explanations

symbols and formatting elements typically found in programming or mathematical expressions

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

file paths and code elements

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

новништво

-0.69

findpost

-0.66

IntoConstraints

-0.65

 متعلقه

-0.63

principalTable

-0.61

 queſto

-0.60

RectangleBorder

-0.59

RenderAtEndOf

-0.59

irten

-0.59



-0.59

POSITIVE LOGITS

пря

0.26

 caméra

0.25

 󰀄

0.24

Ehr

0.24

Mak

0.24

 salarios

0.23

mAdapter

0.23

Sti

0.23

 Dani

0.23

 jurado

0.23

Activations Density 0.337%

symbols and formatting elements typically found in programming or mathematical expressions

file paths and code elements

No Comments

No Known Activations