INDEX

Explanations

references to sequential organization or lists of items

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 7-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L7R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.7.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

Mug

-0.15

wc

-0.14

ought

-0.14

orners

-0.14

Dag

-0.14

ORB

-0.13

boz

-0.13

/sites

-0.13

lined

-0.13

POSITIVE LOGITS

pil

0.15

ãĥ¬ãĥ¼

0.14

asure

0.14

aped

0.14

aping

0.13

ĺ

0.13

nnen

0.13

_LAYOUT

0.13

UA

0.13

emann

0.13

Activations Density 0.045%

references to sequential organization or lists of items

No Comments

No Known Activations