INDEX

Explanations

proper nouns and their associations in a narrative context

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

names and titles

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

 nhau

-0.45

 Warbler

-0.42

HasForeignKey

-0.42

 Observation

-0.40

Wh

-0.40

 temptation

-0.39

URN

-0.39

amaged

-0.38

sim

-0.38

bayaran

-0.38

POSITIVE LOGITS

Personendaten

0.98

AddTagHelper

0.65

rungsseite

0.60

AndEndTag

0.58

 مشين

0.57

ngdoc

0.54

 founder

0.52

 fondateur

0.51

Referanser

0.47

homonymie

0.47

Activations Density 0.051%

proper nouns and their associations in a narrative context

names and titles

No Comments

No Known Activations