INDEX

Explanations

programming constructs related to lambda expressions and method references

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

ivil

-0.15

 civil

-0.14

efined

-0.14

mens

-0.13

ir

-0.13

ç

-0.13

 obstacle

-0.13

 Shorts

-0.13

del

-0.13

.State

-0.13

POSITIVE LOGITS

":[{↵

0.16

_POLICY

0.16

Ctrls

0.15

Î¸ÎµÎ½

0.15

 èī

0.15

fila

0.14

.handler

0.14

Ø±Ø²

0.14

ifter

0.14

 ê²ģ

0.14

Activations Density 0.097%

programming constructs related to lambda expressions and method references

No Comments

No Known Activations