INDEX

Explanations

references to military defense strategies and systems

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

New Auto-Interp

Configuration

fnlp/Llama3_1-8B-Base-LXM-8x/Llama3_1-8B-Base-L22M-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.22.hook_mlp_out

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

inx

-0.16

 Manual

-0.15

_FINE

-0.14

 éĥ

-0.14

amm

-0.14

 manual

-0.14

 byteArray

-0.14

 manuals

-0.14

uren

-0.14

ding

-0.14

POSITIVE LOGITS

pul

0.17

hoff

0.16

hunt

0.16

 Pett

0.15

pon

0.15

cot

0.15

hatt

0.15

rouw

0.15

hattan

0.14

hw

0.14

Activations Density 0.090%

references to military defense strategies and systems

No Comments

No Known Activations