INDEX

Explanations

references to increases or occurrences related to spikes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

imers

-0.17

 persistent

-0.15

orraine

-0.15

rip

-0.14

ners

-0.14

Ú¯Ø§ÙĨÛĮ

-0.14

opa

-0.14

ner

-0.14

wie

-0.14

ird

-0.14

POSITIVE LOGITS

amac

0.17

otic

0.15

thora

0.14

arte

0.14

arts

0.14

isel

0.14

utta

0.14

Walker

0.13

fen

0.13

Activations Density 0.002%

references to increases or occurrences related to spikes

No Comments

No Known Activations