INDEX

Explanations

references to procedural or instructional content related to safety and risk management

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

Ø§ÙģØ±

-0.15

@js

-0.15

anship

-0.14

Ø®Ø´

-0.14

 Beled

-0.14

 SPDX

-0.14

 Townsend

-0.14

 Publication

-0.14

gba

-0.13

âĸį

-0.13

POSITIVE LOGITS

 sort

0.16

Usa

0.16

jud

0.16

 sorts

0.16

Jud

0.15

Ply

0.14

_latency

0.14

Activations Density 0.187%

references to procedural or instructional content related to safety and risk management

No Comments

No Known Activations