INDEX

Explanations

phrases related to gathering information or offering insights

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

ÏģÏİÎ½

-0.14

prog

-0.14

IBE

-0.14

 HERE

-0.14

 ActionTypes

-0.13

 ALSO

-0.13

ragments

-0.13

 actionTypes

-0.13

Å¥

-0.13

_REF

-0.13

POSITIVE LOGITS

 tons

0.50

 loads

0.47

 lots

0.42

 LOAD

0.38

 plenty

0.37

so

0.37

 tonnes

0.37

 quite

0.36

 Loads

0.35

 enough

0.29

Activations Density 0.988%

phrases related to gathering information or offering insights

No Comments

No Known Activations