INDEX

Explanations

references to figures and tables in scientific literature

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

figure and table references

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

uns

-0.44

-0.39

hiz

-0.39

sap

-0.38

part

-0.36

sap

-0.35

const

-0.35

<eos>

-0.34

>>

-0.34

POSITIVE LOGITS

 propOrder

0.79

 ویکی‌پدیا

0.73

 ModelExpression

0.60

principalColumn

0.60

ChildScrollView

0.59

rrggbb

0.58

ViewInit

0.57

Kjelder

0.57

 متعلقه

0.57

 Vikipedi

0.55

Activations Density 0.133%

references to figures and tables in scientific literature

figure and table references

No Comments

No Known Activations