INDEX

Explanations

references to a particular subject or item related to "the."

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

Code, math or bracketed text

np_acts-logits-general · gemini-2.0-flash

math notation and code fragments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

IntoConstraints

-1.03

ंदीखरीदारी

-1.02

twimg

-1.00

tvguidetime

-0.97

awtextra

-0.97

httphttps

-0.95

IsContent

-0.93

 unknownFields

-0.92

MigrationBuilder

-0.91

 שוליים

-0.91

POSITIVE LOGITS

ul

0.56

rul

0.43

UL

0.41

 grounds

0.41

от

0.39

üße

0.39

rile

0.39

entia

0.38

juk

0.38

oul

0.37

Activations Density 0.094%

references to a particular subject or item related to "the."

Code, math or bracketed text

math notation and code fragments

No Comments

No Known Activations