INDEX

Explanations

names of individuals, particularly with the first name "Thomas" or "Tom."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

ouz

-0.16

ToLocal

-0.15

enta

-0.14

ynes

-0.14

ipse

-0.14

elyn

-0.14

obao

-0.14

idlo

-0.14

nds

-0.13

cbc

-0.13

POSITIVE LOGITS

-chan

0.16

ors

0.15

uke

0.15

Ð½Ð°Ð²

0.15

er

0.15

zilla

0.14

uto

0.14

echa

0.14

egal

0.14

 å±

0.14

Activations Density 0.019%

names of individuals, particularly with the first name "Thomas" or "Tom."

No Comments

No Known Activations