INDEX

Explanations

expressions of personal reflection and opinion

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

contingent on subsequent clauses

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

SourceChecksum

-0.68

MigrationBuilder

-0.66

 snippetHide

-0.61

EDEFAULT

-0.58

 незавершена

-0.57

ViewImports

-0.55

 cref

-0.53

enderror

-0.52

 resourceCulture

-0.49

цездатний

-0.49

POSITIVE LOGITS

attach

0.35

limit

0.35

Cat

0.34

 aDecoder

0.34

limits

0.33

fair

0.33

hope

0.33

ChrTalk

0.32

blah

0.32

cat

0.31

Activations Density 0.063%

expressions of personal reflection and opinion

contingent on subsequent clauses

No Comments

No Known Activations