INDEX

Explanations

occurrences of brackets or bracketed text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

vida

-0.16

anson

-0.16

lla

-0.15

Ú©Ø§Øª

-0.15

CADE

-0.14

essen

-0.14

wright

-0.14

ll

-0.14

ki

-0.14

_TURN

-0.14

POSITIVE LOGITS

lus

0.15

rage

0.14

afen

0.14

.jasper

0.14

Ã¤ge

0.14

 ÙĦØ±

0.13

 Ingen

0.13

BASH

0.13

alent

0.13

 Isle

0.13

Activations Density 0.010%

occurrences of brackets or bracketed text

No Comments

No Known Activations