INDEX

Explanations

references to personal pronouns or reflexive pronouns

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_2/width_16k/average_l0_141

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

 purpoſe

-1.32

 Majefty

-1.26

 faſt

-1.21

 houſe

-1.18

 raiſ

-1.17

 pleaſure

-1.14

 ſtate

-1.12

 Anſ

-1.12

 Perſ

-1.11

 Houſe

-1.10

POSITIVE LOGITS

se

1.33

Se

0.93

si

0.90

es

0.87

le

0.77

be

0.77

je

0.74

../../

0.72

0.71

ab

0.71

Activations Density 0.028%

references to personal pronouns or reflexive pronouns

No Comments

No Known Activations