INDEX

Explanations

programming-related keywords, specifically constants and variables in code structures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.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

setVerticalGroup

-0.64

 BoxFit

-0.64

IUrlHelper

-0.62

ContentAlignment

-0.62

principalTable

-0.61

complexContent

-0.61

 makeStyles

-0.59

EndContext

-0.59

AntiForgeryToken

-0.59

 gainera

-0.58

POSITIVE LOGITS

 purpoſe

0.90

 reaſon

0.76

 abſ

0.75

 Anſ

0.75

 cauſe

0.75

 pleaſure

0.73

 ſay

0.73

 ſtate

0.73

 deſt

0.72

agascar

0.70

Activations Density 0.011%

programming-related keywords, specifically constants and variables in code structures

No Comments

No Known Activations