INDEX

Explanations

references to specific rankings or awards associated with the letter "D"

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

iram

-0.19

jvu

-0.19

raw

-0.18

esc

-0.17

rop

-0.17

ocs

-0.17

ropp

-0.17

ictionary

-0.16

ays

-0.16

anny

-0.16

POSITIVE LOGITS

apo

0.16

coded

0.15

ael

0.15

 Roose

0.15

 prox

0.15

anela

0.15

-REAL

0.15

Ñħ

0.15

lad

0.14

lh

0.14

Activations Density 0.035%

references to specific rankings or awards associated with the letter "D"

No Comments

No Known Activations