INDEX

Explanations

questions and inquiries typically starting with "What"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

bow

-0.18

¹Ħ

-0.15

arken

-0.15

#af

-0.15

_acquire

-0.14

 Irvine

-0.14

Ð²Ð¸Ñī

-0.14

usive

-0.14

SelectionMode

-0.14

Ð°Ð»Ð¾Ð²

-0.14

POSITIVE LOGITS

 Does

0.21

Is

0.18

 does

0.18

Is

0.18

Does

0.17

Are

0.17

did

0.16

 Would

0.16

_does

0.15

agraph

0.15

Activations Density 0.055%

questions and inquiries typically starting with "What"

No Comments

No Known Activations