INDEX

Explanations

specific details or planned actions within a context

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

were performed

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

top

-0.45

max

-0.42

 brutal

-0.42

Riv

-0.41

 marginal

-0.40

sub

-0.40

 powerful

-0.40

 small

-0.40

ByType

-0.38

dis

-0.38

POSITIVE LOGITS

 arrangements

0.87

Arrangements

0.86

 Arrangements

0.85

arrange

0.69

MLLoader

0.69

inerary

0.68

 ARRANG

0.68

arrangement

0.68

 Arrangement

0.68

 arrangement

0.68

Activations Density 0.000%

specific details or planned actions within a context

were performed

No Comments

No Known Activations