INDEX

Explanations

various symbols and punctuation typically used in programming syntax

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

closing parenthesis followed by curly brace

np_acts-logits-general · gemini-2.5-flash-lite

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

 <<<<<<<<<<<<<<

-0.59

fromnode

-0.57

帖最后由

-0.56

]")]

-0.55

__*/

-0.53

ValueStyle

-0.52

 Grüsse

-0.49

rrggbb

-0.48

jsonwebtoken

-0.48

 INTERESAR

-0.48

POSITIVE LOGITS

ModelAttribute

0.41

 Roskov

0.40

 Савезне

0.39

RequestBody

0.38

ThroughAttribute

0.37

vocato

0.35

 patient

0.35

 Patient

0.35

 msglen

0.33

 ContentValues

0.32

Activations Density 0.004%

various symbols and punctuation typically used in programming syntax

closing parenthesis followed by curly brace

No Comments

No Known Activations