INDEX

Explanations

vague and indeterminate references to concepts and items

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ynn

-0.15

lo

-0.14

 Conway

-0.14

cona

-0.14

ï¸

-0.14

esc

-0.13

MatrixMode

-0.13

Ã

-0.13

ulary

-0.13

/cpu

-0.13

POSITIVE LOGITS

vae

0.15

ash

0.15

ÄįÃŃ

0.15

mq

0.13

 sund

0.13

ÏģÎ¹Î½

0.13

erne

0.13

çķª

0.13

Î·

0.13

nings

0.13

Activations Density 0.024%

vague and indeterminate references to concepts and items

No Comments

No Known Activations