INDEX

Explanations

phrases that describe precision or quality in various contexts, often emphasizing "fine" details or classifications

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

"fine"

np_acts-logits-general · gemini-2.0-flash

fine as a modifier

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

yntaxException

-0.59

rrggbb

-0.54

Cyfeiriadau

-0.53

ickness

-0.51

СР

-0.48

richtig

-0.48

tvguidetime

-0.48

ypsies

-0.47

rikes

-0.47

 cesar

-0.47

POSITIVE LOGITS

 FINE

1.12

fine

1.10

FINE

1.09

Fine

1.09

 grained

1.09

 tuned

1.08

 tuning

1.07

 Fine

1.05

 fine

1.04

 Fines

1.00

Activations Density 0.065%

phrases that describe precision or quality in various contexts, often emphasizing "fine" details or classifications

"fine"

fine as a modifier

No Comments

No Known Activations