INDEX

Explanations

recurring phrases centered around the word "to" and conjunctions like "and."

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

connectors and conjunctions

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

Diweddarwch

-0.39

 hvil

-0.33

 unfavorable

-0.33

ousness

-0.33

ousands

-0.33

 reca

-0.32

和田

-0.32

ともに

-0.32

alyptus

-0.31

 kaynağından

-0.31

POSITIVE LOGITS

 odkazy

0.58

VolleyError

0.56

RTEE

0.55

GOTREF

0.54

SequentialGroup

0.54

+#+#

0.54

 perchè

0.53

.*")]

0.53

 poichè

0.52

 propOrder

0.52

Activations Density 0.128%

recurring phrases centered around the word "to" and conjunctions like "and."

connectors and conjunctions

No Comments

No Known Activations