INDEX

Explanations

code-related elements, particularly identifiers and function calls

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

FormTagHelper

-0.67



-0.64

GINS

-0.48

hoga

-0.47

AntiForgeryToken

-0.46

StartTag

-0.46

principalTable

-0.45

bustible

-0.45

askins

-0.44

NewGuid

-0.43

POSITIVE LOGITS

 myſelf

0.83

 purpoſe

0.78

 pleaſure

0.76

 Majefty

0.75

 itſelf

0.72

 houſe

0.71

 themſelves

0.71

 himſelf

0.70

neſs

0.70

ſelves

0.69

Activations Density 0.090%

code-related elements, particularly identifiers and function calls

No Comments

No Known Activations