INDEX

Explanations

references to specific roles, titles, and structures associated with organization or hierarchy

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

"top" followed by other words

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

Organisateur

-0.47

Cyfarwyddwr

-0.38

agie

-0.36

 previous

-0.36

 extensive

-0.35

 Grimm

-0.35

[]=

-0.34

Har

-0.34

setVerticalGroup

-0.33

[]

-0.33

POSITIVE LOGITS

Top

1.16

top

1.16

Top

1.13

getTop

1.01

top

1.01

TOP

1.01

 Tops

0.97

 tops

0.94

TOP

0.93

トップ

0.91

Activations Density 0.021%

references to specific roles, titles, and structures associated with organization or hierarchy

"top" followed by other words

No Comments

No Known Activations