INDEX

Explanations

occurrences of the word "exec" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_0/width_16k/average_l0_105

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

-0.65

He

-0.57

<eos>

-0.56

the

-0.55

di

-0.54

vertret

-0.53

al

-0.51

CDCl

-0.51

ricus

-0.51

POSITIVE LOGITS

 myſelf

1.13

 ſtate

1.12

 purpoſe

1.06

 reaſon

1.03

 houſe

1.03

 fubject

1.02

 ſche

1.01

ſelf

1.01

 pleaſure

1.01

 Majefty

1.00

Activations Density 0.113%

occurrences of the word "exec" in various contexts

No Comments

No Known Activations