INDEX

Explanations

code-related terminologies and structures, particularly those related to data processing and testing frameworks

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

Code-related syntax and characters

np_acts-logits-general · gemini-2.0-flash

URLs and code snippets

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

 "..\..\

-0.61

chap

-0.50

 "..\..\..\

-0.49

TagMode

-0.49

 /\.(

-0.48

usepackage

-0.48

enumii

-0.45

 словом

-0.45

edos

-0.45

guer

-0.44

POSITIVE LOGITS

 absl

0.60

 Yvette

0.56

DCHECK

0.55

 compositor

0.55

 població

0.55

 stället

0.54

 Albuquerque

0.53

 gyhoeddwyd

0.52

 fuchsia

0.51

gdala

0.51

Activations Density 0.377%

code-related terminologies and structures, particularly those related to data processing and testing frameworks

Code-related syntax and characters

URLs and code snippets

No Comments

No Known Activations