INDEX

Explanations

references to performance and functionality in various contexts

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

worked wonderfully

np_acts-logits-general · gemini-2.5-flash-lite

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

orianCalendar

-0.40

jaros

-0.37

ellschaft

-0.35

 Downvote

-0.35

igrette

-0.35

 want

-0.35

 FetchType

-0.34

 LIABLE

-0.34

regard

-0.33

 Anglia

-0.33

POSITIVE LOGITS

 excellently

0.60

 dramatically

0.57

IBOutlet

0.56

 impressively

0.55

 superbly

0.55

 ExecuteAsync

0.54

 wonderfully

0.54

 amazingly

0.52

 excellent

0.52

ükemmel

0.52

Activations Density 0.054%

references to performance and functionality in various contexts

worked wonderfully

No Comments

No Known Activations