INDEX

Explanations

phrases indicating ongoing actions or progress, particularly those involving the word "on"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

Personendaten

-1.04

 myſelf

-0.88

 themſelves

-0.87

 Reſ

-0.87

 itſelf

-0.83

 pleaſure

-0.75

complexContent

-0.75

 againſt

-0.74

noons

-0.74

 leaſt

-0.74

POSITIVE LOGITS

 track

0.68

 board

0.65

 mission

0.62

 task

0.56

 alert

0.53

mission

0.52

 Board

0.50

 verge

0.48

task

0.47

 Mission

0.46

Activations Density 0.089%

phrases indicating ongoing actions or progress, particularly those involving the word "on"

No Comments

No Known Activations