INDEX

Explanations

conjunctions and definite articles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

amik

-0.15

 îł

-0.15

illis

-0.15

strup

-0.15

appa

-0.14

 Airways

-0.14

ROSS

-0.14

stown

-0.14

orda

-0.14

quee

-0.14

POSITIVE LOGITS

zes

0.18

0.16

leck

0.15

ingle

0.15

SEC

0.15

/animations

0.14

Ped

0.14

sec

0.13

cce

0.13

ÙĦØ¨

0.13

Activations Density 0.005%

conjunctions and definite articles

No Comments

No Known Activations