INDEX

Explanations

phrases indicating presence and qualities of entities or subjects

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

Kathy, kinase, hotel, climate, relationship

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

 препратки

-0.52

 beginnetje

-0.50

 ProtoMessage

-0.47

 typelib

-0.47

 Italij

-0.46

 alrededores

-0.45

MigrationBuilder

-0.44

actéristi

-0.44

 dieß

-0.43

 prohibido

-0.43

POSITIVE LOGITS

первых

0.53

setVerticalGroup

0.53

xer

0.51

ulico

0.49

fillType

0.49

͝

0.49

 XCTest

0.49

getColumnIndex

0.49

RCT

0.49

 Brid

0.48

Activations Density 0.275%

phrases indicating presence and qualities of entities or subjects

Kathy, kinase, hotel, climate, relationship

No Comments

No Known Activations