INDEX

Explanations

references to hats and headwear

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

 bufsize

-0.16

 shirts

-0.16

inia

-0.16

 dresses

-0.15

dle

-0.14

 Shirt

-0.14

ecure

-0.14

endon

-0.14

shirt

-0.14

POSITIVE LOGITS

hat

0.60

 hats

0.53

hat

0.53

Hat

0.51

Hat

0.48

å¸½

0.47

 Hats

0.42

cap

0.41

_hat

0.41

 caps

0.40

Activations Density 0.089%

references to hats and headwear

No Comments

No Known Activations