INDEX

Explanations

error messages and warnings in code

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

Punctuation, often quotes or parentheses

np_acts-logits-general · gemini-2.0-flash

error or invalid messages

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

-0.47

ObjectMeta

-0.41

hot

-0.39

Mutable

-0.38

poly

-0.38

ahu

-0.36

 poly

-0.36

setAccessible

-0.35

 varm

-0.35

ich

-0.35

POSITIVE LOGITS

 propOrder

0.92

 itſelf

0.92

ſelf

0.91

 invalid

0.89

 insufficient

0.88

 reaſon

0.87

 Invalid

0.87

 wrong

0.86

 Jefus

0.86

 Monfieur

0.86

Activations Density 0.162%

error messages and warnings in code

Punctuation, often quotes or parentheses

error or invalid messages

No Comments

No Known Activations