INDEX

Explanations

code-related keywords and constructs

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 35-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_35/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.35.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

oa̍t

-0.49

-------

-0.41

providedIn

-0.38

 الطبي

-0.36

Égypte

-0.35

 ücret

-0.35

 virgen

-0.35

 soins

-0.35

 Hertfordshire

-0.35

richtet

-0.34

POSITIVE LOGITS

 iprot

0.68

thrift

0.66

 Thrift

0.64

 thrift

0.64

 oprot

0.61

Thrift

0.58

 apache

0.54

writeFieldEnd

0.53

 Apache

0.53

querque

0.53

Activations Density 0.312%

code-related keywords and constructs

No Comments

No Known Activations