INDEX

Explanations

references to alignment or coherence in various contexts

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

ãĤ¥

-0.18

Ø®Ø§ÙĨÙĩ

-0.15

stown

-0.15

lyn

-0.15

elyn

-0.14

illy

-0.14

-eslint

-0.14

erm

-0.14

idal

-0.14

inston

-0.14

POSITIVE LOGITS

arity

0.18

 perfectly

0.17

ean

0.17

-sama

0.17

 forces

0.17

 interests

0.17

EMENT

0.17

uju

0.17

 efforts

0.16

erland

0.16

Activations Density 0.051%

references to alignment or coherence in various contexts

No Comments

No Known Activations