INDEX

Explanations

references to result variables and data structures in programming contexts

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

Code snippets or program output

np_acts-logits-general · gemini-2.0-flash

code tokens and programming constructs

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 referrerpolicy

-1.00

########.

-0.85

AddTagHelper

-0.82

CppMethod

-0.79

IsContent

-0.77

stateProvider

-0.77

 esternos

-0.73

basicConfig

-0.73

 оригіналу

-0.73

 незавершена

-0.73

POSITIVE LOGITS

 such

0.25

 bedenken

0.25

 instantly

0.25

 respecto

0.24

 cent

0.23

 nedost

0.22

©

0.22

 corn

0.22

 static

0.22

th

0.21

Activations Density 0.397%

references to result variables and data structures in programming contexts

Code snippets or program output

code tokens and programming constructs

No Comments

No Known Activations