INDEX

Explanations

statements and instructions relevant to legal or medical contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_22/width_16k/average_l0_123

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

IUrlHelper

-0.33

 living

-0.29

 Entrepreneurs

-0.28

 experimentation

-0.28

 experiment

-0.28

 entrepreneurs

-0.26

 undist

-0.26

 theoretically

-0.25

 experimental

-0.25

int

-0.25

POSITIVE LOGITS

 wording

0.87

 worded

0.86

 phrases

0.81

 nakalista

0.81

 phrase

0.73

 frasi

0.73

 Phrases

0.73

phrases

0.71

 phrasing

0.70

phrase

0.70

Activations Density 0.656%

statements and instructions relevant to legal or medical contexts

No Comments

No Known Activations