INDEX

Explanations

the letter B, which suggests it may be identifying references, or possibly gene names

oai_token-act-pair · gemini-2.0-flash

code identifiers or syntax patterns

np_token-act-pair-logits · gpt-4o-mini

B

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_21/width_16k/average_l0_13

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.ln2.hook_normalized

Architecture

jumprelu_transcoder

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 read

-0.49

 Amber

-0.45

 READ

-0.45

read

-0.43

thâu

-0.43

testcase

-0.43

Amber

-0.42

 master

-0.42

Smith

-0.41

 κατά

-0.41

POSITIVE LOGITS

 Chwiliwch

0.71

zzleHttp

0.70

ScopeManager

0.70

 وتسجيلات

0.63

$_['

0.61

 AssemblyCulture

0.60

DTD

0.59

/\.

0.59

Šaltiniai

0.58

etheless

0.57

Activations Density 0.163%

the letter B, which suggests it may be identifying references, or possibly gene names

code identifiers or syntax patterns

B

No Comments

No Known Activations