INDEX

Explanations

references to poisoning or toxic substances

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

DebuggerNonUser

-0.61

 glyphicon

-0.60

 AssertionError

-0.58

 typelib

-0.55

 $(\%)$

-0.55

 shutil

-0.50

rån

-0.49

="#"><

-0.49

 Geschlecht

-0.48

 RATE

-0.47

POSITIVE LOGITS

 poison

1.44

poison

1.35

 Poison

1.33

Poison

1.31

 poisoning

1.28

 poisoned

1.27

 poisons

1.26

 poisonous

1.21

Toxic

1.05

 toxic

1.05

Activations Density 1.441%

references to poisoning or toxic substances

No Comments

No Known Activations