INDEX

Explanations

names and titles related to authors or media figures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

luetooth

-0.17

sul

-0.15

ä¾¯

-0.14

 tant

-0.14

 Disp

-0.14

 queryInterface

-0.14

ouri

-0.14

sal

-0.13

AMIL

-0.13

thon

-0.13

POSITIVE LOGITS

Tup

0.17

Łèĥ½

0.16

ipa

0.15

iem

0.14

vocab

0.14

tap

0.14

fle

0.14

inski

0.14

Tap

0.14

osu

0.14

Activations Density 0.107%

names and titles related to authors or media figures

No Comments

No Known Activations