INDEX

Explanations

programming constructs and keywords related to object manipulation in a coding context

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 41-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_41/width_16k/average_l0_113

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.41.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

MessageTagHelper

-0.46

RectangleBorder

-0.44

BagLayout

-0.43

 draught

-0.43

 ویکی‌پدی

-0.42

ftagPool

-0.41

 Signalez

-0.40

 snippetHide

-0.40

RTGC

-0.39

 pylori

-0.39

POSITIVE LOGITS

 Quake

0.56

 Valve

0.54

 client

0.50

 Gordon

0.50

ent

0.49

edict

0.48

 Client

0.48

clients

0.48

WriteByte

0.47

client

0.47

Activations Density 0.194%

programming constructs and keywords related to object manipulation in a coding context

No Known Activations