INDEX

Explanations

contrasts or comparisons within the text

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

arb

-0.15

Åĵ

-0.15

-0.14

ï¼ĸ

-0.14

allet

-0.14

 perman

-0.14

eneg

-0.14

 altern

-0.14

mock

-0.14

-0.13

POSITIVE LOGITS

åºľ

0.15

anto

0.15

awy

0.14

 Flesh

0.14

ypad

0.14

ÑħÐ¾Ð´

0.14

sst

0.13

è«ĸ

0.13

.idea

0.13

.food

0.13

Activations Density 0.102%

contrasts or comparisons within the text

No Comments

No Known Activations