INDEX

Explanations

titles and names of exhibitions or events

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

 Picker

-0.16

 Thompson

-0.14

actly

-0.14

 bore

-0.14

pcm

-0.13

ÙĪØ±Ø§ÙĨ

-0.13

 vice

-0.13

Åº

-0.13

ãĥ¼ãĥĢ

-0.13

 breathing

-0.13

POSITIVE LOGITS

icolor

0.15

itta

0.15

itra

0.15

uler

0.14

prite

0.14

iza

0.14

.scalar

0.14

chaft

0.14

itel

0.13

nit

0.13

Activations Density 0.038%

titles and names of exhibitions or events

No Comments

No Known Activations

titles and names of exhibitions or events

No Comments

No Known Activations