INDEX

Explanations

expressions of self-reference

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L1R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.1.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

dera

-0.20

tone

-0.16

chez

-0.16

reesome

-0.16

wort

-0.15

alet

-0.14

uche

-0.14

akit

-0.14

ÑĤÑĮÑģÑı

-0.14

ilot

-0.14

POSITIVE LOGITS

 belt

0.16

ycop

0.15

 daytime

0.15

 wish

0.15

 Schl

0.14

ileged

0.13

 recip

0.13

ÑĨÐ¸Ð¿

0.13

 Char

0.13

Activations Density 0.086%

expressions of self-reference

No Comments

No Known Activations