INDEX

Explanations

programming languages and code-related terms

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

bootstrapcdn

-0.56

DIPSETTING

-0.51

 betweenstory

-0.44

principalColumn

-0.41

RegressionTest

-0.38

🔕

-0.36

BSITE

-0.36

IActionResult

-0.36

libft

-0.36

Fprintf

-0.35

POSITIVE LOGITS

 snippetHide

0.93

хьтан

0.57

astify

0.54

 nakalista

0.54

zzlement

0.48

stak

0.47

 مشين

0.46

 NSFW

0.46

 iconFacebook

0.46

jsxFileName

0.46

Activations Density 0.074%

programming languages and code-related terms

No Comments

No Known Activations