INDEX

Explanations

key elements or concepts relevant to a specific narrative or context

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

脚注の使い方

-0.33

期刊论文

-0.32

SR

-0.32

tonode

-0.30

 Better

-0.30

rehensive

-0.29

 różnych

-0.29

做一个

-0.29

 scant

-0.29

 Recursos

-0.28

POSITIVE LOGITS

 hardest

0.62

 closest

0.61

 furthest

0.59

 najbardziej

0.57

 createSlice

0.56

CppMethod

0.56

 farthest

0.56

 saddest

0.54

 slowest

0.53

tableFuture

0.52

Activations Density 0.061%

key elements or concepts relevant to a specific narrative or context

No Comments

No Known Activations