INDEX

Explanations

phrases that reference lists or structured formats, particularly in relation to "line" items

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

ness

-0.21

mol

-0.17

name

-0.17

nt

-0.17

teen

-0.16

lick

-0.16

marine

-0.16

lings

-0.16

ly

-0.16

imb

-0.16

POSITIVE LOGITS

arity

0.46

aments

0.37

ament

0.35

ups

0.33

ages

0.28

age

0.25

haul

0.25

amenti

0.25

-up

0.25

amientos

0.24

Activations Density 0.080%

phrases that reference lists or structured formats, particularly in relation to "line" items

No Comments

No Known Activations