INDEX

Explanations

the name "Bernard" or related variants in various contexts

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

MOOTH

-0.18

bet

-0.18

amp

-0.18

age

-0.17

ermann

-0.16

/xml

-0.15

eacher

-0.15

ties

-0.15

itan

-0.15

ched

-0.14

POSITIVE LOGITS

rd

0.19

================================================================

0.18

yth

0.17

(es

0.16

------------------------------------------------------------------------------------------------

0.15

optera

0.15

enthal

0.14

infile

0.14

ese

0.14

eg

0.14

Activations Density 0.290%

the name "Bernard" or related variants in various contexts

No Comments

No Known Activations