INDEX

Explanations

phrases concerning fairness and moral judgments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Aim

-0.16

jak

-0.15

 Pret

-0.15

uis

-0.14

Wet

-0.14

 pret

-0.13

iso

-0.13

pron

-0.13

orgen

-0.13

aver

-0.13

POSITIVE LOGITS

 fair

0.42

 proper

0.40

 appropriate

0.40

appropriate

0.40

fair

0.35

 correct

0.33

 fairness

0.33

 Fair

0.32

Fair

0.30

 right

0.30

Activations Density 0.188%

phrases concerning fairness and moral judgments

No Comments

No Known Activations