INDEX

Explanations

references to information or data structures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 3-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_3/width_16k/average_l0_90

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.3.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

PYX

-0.42

GenerationType

-0.42

 cellpadding

-0.41

waitKey

-0.38

HtmlAttribute

-0.36

 comuniques

-0.36

jspb

-0.35

HomeAsUpEnabled

-0.35

पया

-0.33

getAdapter

-0.33

POSITIVE LOGITS

Info

2.52

 Info

1.97

 INFO

1.17

UserInfo

1.12

Infos

1.12

 Information

1.10

 Infos

1.10

INFO

1.08

 UserInfo

1.04

ItemInfo

1.02

Activations Density 0.120%

references to information or data structures

No Comments

No Known Activations