INDEX

Explanations

concepts related to goals, capabilities, and intentions

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

 CircularProgress

-0.18

inne

-0.16

ienie

-0.15

duk

-0.14

egas

-0.14

zug

-0.14

anik

-0.14

 buurt

-0.14

ulsive

-0.14

PV

-0.13

POSITIVE LOGITS

ks

0.15

ofs

0.15

 delivering

0.14

 tore

0.14

_SI

0.14

 being

0.14

 mass

0.14

to

0.14

_kwargs

0.14

Activations Density 0.102%

concepts related to goals, capabilities, and intentions

No Comments

No Known Activations