INDEX

Explanations

phrases indicating evaluations and comparisons

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

aber

-0.17

 schemes

-0.16

ango

-0.16

èĨľ

-0.15

mum

-0.15

å¼ķãģį

-0.15

Scheme

-0.15

hd

-0.15

 Trie

-0.15

 scheme

-0.15

POSITIVE LOGITS

cos

0.15

elta

0.15

oft

0.15

analyze

0.15

Anc

0.15

ar

0.15

issen

0.15

èĬ³

0.14

sod

0.14

dess

0.14

Activations Density 0.458%

phrases indicating evaluations and comparisons

No Comments

No Known Activations