INDEX

Explanations

negations or commands regarding actions in a variety of contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

usher

-0.16

ansk

-0.15

urd

-0.14

Ð±Ð°Ð²

-0.14

_slots

-0.14

ntax

-0.14

pag

-0.13

ãĥ¼ãĥ¬

-0.13

apiro

-0.13

ansen

-0.13

POSITIVE LOGITS

 forget

0.17

 trust

0.17

tempt

0.17

æĢ¥

0.16

 touching

0.16

 ever

0.15

rush

0.15

 reliance

0.15

 EVER

0.15

 expect

0.15

Activations Density 0.144%

negations or commands regarding actions in a variety of contexts

No Comments

No Known Activations