INDEX

Explanations

sequences of underscores or whitespace characters

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

Variables ending in "perf"

np_acts-logits-general · gemini-2.0-flash

variable assignments

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

-0.79

–

-0.77

-0.72

—

-0.69

-0.66

His

-0.65

-0.64

Tr

-0.63

-0.62

POSITIVE LOGITS

 myſelf

1.39

 itſelf

1.25

 greateſt

1.20

 pleaſure

1.19

 Roskov

1.17

 ſche

1.13

 doubtnut

1.12

 crdi

1.10

 Monfieur

1.10

 ſtate

1.09

Activations Density 0.683%

sequences of underscores or whitespace characters

Variables ending in "perf"

variable assignments

No Comments

No Known Activations