INDEX

Explanations

punctuation, particularly commas, indicating pauses or separations in thoughts

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

following conjunctions and transitions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

New

-0.34

-0.33

Path

-0.33

Track

-0.33

based

-0.32

where

-0.32

Con

-0.31

POSITIVE LOGITS

 oprot

0.83

 चीज़ों

0.81

 CreateTagHelper

0.76

 Chwiliwch

0.75

 dezelve

0.75

 transfieras

0.73

 ſind

0.73

 autorytatywna

0.72

$_(

0.72

rungsseite

0.72

Activations Density 0.045%

punctuation, particularly commas, indicating pauses or separations in thoughts

following conjunctions and transitions

No Comments

No Known Activations