INDEX

Explanations

suffix patterns related to characteristics or descriptions

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

CHANT

-0.16

ooter

-0.16

 Å¾al

-0.16

asti

-0.15

lever

-0.15

kins

-0.15

_REPEAT

-0.15

ĥ

-0.15

fone

-0.14

DNA

-0.14

POSITIVE LOGITS

Han

0.16

0.15

acular

0.15

 Georgia

0.14

 éĢ

0.14

Activations Density 0.001%

suffix patterns related to characteristics or descriptions

No Comments

No Known Activations