INDEX

Explanations

terminology related to legal and ethical issues, particularly focusing on accusations and judgments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

 raiſ

-1.14

 ſche

-1.09

 ſever

-1.06

ſelves

-1.06

 myſelf

-1.05

 itſelf

-1.04

 Reſ

-1.03

 pleaſure

-1.01

ſelf

-0.99

 Jefus

-0.98

POSITIVE LOGITS

[toxicity=0]

0.81

</i>

0.65

AnchorStyles

0.63

0.59

of

0.56

󠁿

0.54

iastes

0.54

¿½

0.53

</em>

0.52

Activations Density 0.362%

terminology related to legal and ethical issues, particularly focusing on accusations and judgments

No Comments

No Known Activations