INDEX

Explanations

references to saddles and related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 2-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L2R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.2.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

essaging

-0.16

soever

-0.15

Ì§

-0.15

ajar

-0.15

chy

-0.15

ivers

-0.14

rag

-0.14

eme

-0.14

 saja

-0.14

cka

-0.14

POSITIVE LOGITS

odzi

0.15

mere

0.15

kok

0.14

oldur

0.14

_digest

0.14

urette

0.14

 casts

0.14

íĸ¥

0.14

²

0.13

ILTER

0.13

Activations Density 0.008%

references to saddles and related terms

No Comments

No Known Activations