INDEX

Explanations

expressions related to results and outcomes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

rz

-0.16

mund

-0.15

thing

-0.15

ylland

-0.14

agnost

-0.14

ASSERT

-0.14

INCT

-0.14

_PROVID

-0.14

awai

-0.14

ãĤ®

-0.14

POSITIVE LOGITS

hood

0.16

 efforts

0.15

aby

0.14

-effects

0.14

ispers

0.14

 research

0.14

ultan

0.14

ippo

0.13

 conducted

0.13

Activations Density 0.066%

expressions related to results and outcomes

No Comments

No Known Activations