INDEX

Explanations

words and phrases related to performance metrics and achievements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

atra

-0.15

rade

-0.15

rado

-0.14

Ã¡ÅĻ

-0.14

ÑĢÐ°ÑĤ

-0.14

arda

-0.13

opsis

-0.13

rak

-0.13

resse

-0.13

ingo

-0.12

POSITIVE LOGITS

run

1.35

Run

1.17

 runs

1.16

run

1.13

Run

1.10

-run

1.10

.run

1.08

RUN

1.08

 running

1.06

ran

1.05

Activations Density 0.529%

words and phrases related to performance metrics and achievements

No Comments

No Known Activations