INDEX

Explanations

descriptions or evaluations characterized by the adverb "very"

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

ikh

-0.17

 Repeat

-0.16

 repeat

-0.15

repeat

-0.15

Repeat

-0.15

 mere

-0.14

Ä±rak

-0.14

enheim

-0.14

SSERT

-0.14

 repeating

-0.14

POSITIVE LOGITS

ãĥĥãĤ·ãĥ¥

0.16

Omn

0.16

kle

0.15

aret

0.15

ara

0.15

onen

0.15

urrets

0.15

çľģ

0.14

fad

0.14

ones

0.14

Activations Density 0.045%

descriptions or evaluations characterized by the adverb "very"

No Comments

No Known Activations