INDEX

Explanations

terms related to instructions or additional information in content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.20.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

iyat

-0.15

allee

-0.15

ITE

-0.15

 Ø´Ø¹

-0.14

roduce

-0.14

_rewrite

-0.14

InternalServerError

-0.14

EMPLARY

-0.14

atoire

-0.14

ega

-0.14

POSITIVE LOGITS

 skip

0.79

 Skip

0.72

 skipping

0.69

skip

0.68

 bypass

0.67

 skipped

0.66

 skips

0.66

Skip

0.65

.skip

0.59

 SKIP

0.58

Activations Density 0.256%

terms related to instructions or additional information in content

No Comments

No Known Activations