INDEX

Explanations

discussions about personal experiences and reflections

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

though, hoping, necessary, good

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

 swimming

-0.45

yticks

-0.42

tawesome

-0.42

:✨

-0.41

 zwem

-0.41

ьажоргаш

-0.41

 Swimming

-0.41

erweise

-0.40

 Stacks

-0.40

 valsty

-0.40

POSITIVE LOGITS

tagHelperRunner

0.42

 محفوظة

0.40

 aDecoder

0.38

AppMethodBeat

0.37

jącą

0.37

esModule

0.35

ない

0.35

nesc

0.35

ooc

0.34

 <<<<<<<<<<<<<<

0.34

Activations Density 0.027%

discussions about personal experiences and reflections

though, hoping, necessary, good

No Comments

No Known Activations