INDEX

Explanations

specific names, likely related to individuals or characters in a cultural context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

lÃ½

-0.16

fax

-0.15

eton

-0.15

igne

-0.15

lÃŃÄį

-0.15

igid

-0.15

anian

-0.15

eza

-0.15

omb

-0.14

ieri

-0.14

POSITIVE LOGITS

Ùĩ

0.19

0.18

awai

0.16

es

0.16

els

0.15

sched

0.15

hazi

0.15

su

0.15

Activations Density 0.010%

specific names, likely related to individuals or characters in a cultural context

No Comments

No Known Activations