INDEX

Explanations

references to entities and roles in a collaborative or procedural context

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

-0.57

-0.45

𝑙

-0.41

ি

-0.41

й

-0.36

𝑖

-0.36

𝑠

-0.35

ヨーク

-0.35

 собой

-0.34

𝑟

-0.34

POSITIVE LOGITS

er

1.02

ergic

0.75

erick

0.73

 BoxDecoration

0.71

erals

0.71

 ſche

0.69

cellaneous

0.69

erus

0.68

 Verſ

0.68

 ſta

0.66

Activations Density 0.174%

references to entities and roles in a collaborative or procedural context

No Comments

No Known Activations