INDEX

Explanations

sections of code related to control flow or logic statements

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

Non-English or unusual words/tokens

np_acts-logits-general · gemini-2.0-flash

disallowing leading to outcomes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 strength

-0.56

참고

-0.51

BASEPATH

-0.49

strength

-0.49

fastjson

-0.48

!$

-0.45

 Strength

-0.44

])]

-0.43

)]

-0.42

POSITIVE LOGITS

aarrggbb

0.87

RenderAtEndOf

0.76

MessageOf

0.74

Efq

0.71

Хьажоргаш

0.67

 Wikimedijinoj

0.67

 Theſe

0.66

 Мексичка

0.64

 Tenth

0.64

InitVars

0.63

Activations Density 0.043%

sections of code related to control flow or logic statements

Non-English or unusual words/tokens

disallowing leading to outcomes

No Comments

No Known Activations