INDEX

Explanations

instances of the letter 'a' in various contexts, often within mathematical or analytical descriptions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

PerformLayout

-0.60

 Wikimedijinoj

-0.52

 AssemblyCulture

-0.50

被打

-0.48

UnknownFields

-0.45

 FILLER

-0.45

ագրություններ

-0.44

Limiter

-0.44

 BrowserModule

-0.44

❲

-0.43

POSITIVE LOGITS

ymce

0.63

RTGC

0.54

 useCallback

0.45

ⓑ

0.40

Jereo

0.40

wantErr

0.40

ⓐ

0.40

ewon

0.38

grafi

0.37

genheit

0.37

Activations Density 0.567%

instances of the letter 'a' in various contexts, often within mathematical or analytical descriptions

No Comments

No Known Activations