INDEX

Explanations

specific numeric and textual patterns or formats that likely correspond to structured data in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 16-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L16R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.16.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

dorf

-0.20

coe

-0.16

orer

-0.15

">//

-0.14

occan

-0.14

ãĥªãĥ³

-0.14

blas

-0.14

://%

-0.14

bakan

-0.13

.drawLine

-0.13

POSITIVE LOGITS

tro

0.20

0.17

 mitt

0.17

icher

0.15

hil

0.14

hong

0.14

ers

0.14

pees

0.14

acob

0.14

irm

0.13

Activations Density 0.078%

specific numeric and textual patterns or formats that likely correspond to structured data in various contexts

No Comments

No Known Activations