INDEX

Explanations

ellipses or truncated text indicating omitted content

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-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

 queſta

-0.71

IBarButtonItem

-0.55

 Autorizaciones

-0.54

Gön

-0.54

 ferons

-0.51

ⓧ

-0.49

 enfans

-0.47

inWeight

-0.46

 actionMode

-0.46

ſammen

-0.45

POSITIVE LOGITS

GOTREF

0.58

RemoveField

0.55

DebuggerNonUser

0.47

putAll

0.46

deepEqual

0.45

AddField

0.44

jsxFileName

0.44

...

0.44

ものの

0.43

memcpy

0.43

Activations Density 0.001%

ellipses or truncated text indicating omitted content

No Comments

No Known Activations