INDEX

Explanations

HTML tags and attributes

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ThroughAttribute

-0.47

RequestMapping

-0.45

canestro

-0.44

sidemargin

-0.44

argc

-0.43

Попис

-0.43

pil

-0.42

tinyos

-0.42

 magn

-0.41

 protoimpl

-0.40

POSITIVE LOGITS

 nakalista

0.67

transQ

0.44

ContentAsync

0.44

 Hälfte

0.42

今度は

0.41

 AssemblyCulture

0.41

 corta

0.41

expliquer

0.41

 Jenderal

0.41

 вікісторінку

0.40

Activations Density 0.051%

HTML tags and attributes

No Comments

No Known Activations