INDEX

Explanations

references to specific individuals or examples in a narrative context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

EventManager

-0.17

lags

-0.16

 eigentlich

-0.15

elian

-0.14

/single

-0.14

 preferably

-0.14

increments

-0.14

_parms

-0.14

acons

-0.14

atypes

-0.13

POSITIVE LOGITS

 such

0.25

 example

0.25

 particularly

0.24

 especially

0.21

 particular

0.21

 estimate

0.20

such

0.20

example

0.20

 possibility

0.20

of

0.19

Activations Density 0.053%

references to specific individuals or examples in a narrative context

No Comments

No Known Activations