INDEX

Explanations

words that are related to various aspects of naming or categorizing individuals or entities

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

Tokens with "le" in them

np_acts-logits-general · gemini-2.0-flash

words ending in -ble, -ggle, -pple, -dle, -ttle

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_16k/average_l0_76

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.31.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

으로

-0.58

-0.51

not

-0.48

ように

-0.47

ness

-0.47

rich

-0.42

ref

-0.42

non

-0.42

ів

-0.42

 EconPapers

-0.41

POSITIVE LOGITS

parsedMessage

0.66

RenderAtEndOf

0.57

 pleaſure

0.56

lable

0.56

0.54

ementara

0.54

tle

0.54

 Houſe

0.52

arangay

0.51

 outfile

0.50

Activations Density 0.347%

words that are related to various aspects of naming or categorizing individuals or entities

Tokens with "le" in them

words ending in -ble, -ggle, -pple, -dle, -ttle

No Comments

No Known Activations