INDEX

Explanations

high numerical values associated with significance or maximum potential in various contexts

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

Batten disease

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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



-0.52

reif

-0.52

évalu

-0.52

 bParam

-0.48

ühr

-0.47

Selfer

-0.47

 Fais

-0.46

ouncil

-0.46

Subclass

-0.46

idosis

-0.46

POSITIVE LOGITS

ten

2.50

TEN

1.93

Ten

1.92

Ten

1.89

ten

1.88

TEN

1.82

テン

1.09

tens

1.07

tena

1.06

 tenn

1.05

Activations Density 0.012%

high numerical values associated with significance or maximum potential in various contexts

Batten disease

No Comments

No Known Activations