INDEX

Explanations

phrases that denote inclusion or mention inclusion directly

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

New Auto-Interp

Configuration

Juliushanhanhan/llama-3-8b-it-res/blocks.25.hook_resid_post

Features

65,536

Data Type

float32

Hook Name

blocks.25.hook_resid_post

Hook Layer

Architecture

gated

Context Size

1,024

Dataset

Juliushanhanhan/openwebtext-1b-llama3-tokenized-cxt-1024

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãģĬãĤĬ

-0.21

uel

-0.16

jev

-0.15

our

-0.14

Ð³Ð¾Ð²

-0.14

ãĥĨãĥ«

-0.14

oleon

-0.13

jax

-0.12

ãĥ¬ãĥ³

-0.12

Occurred

-0.12

POSITIVE LOGITS

/ex

0.30

ognito

0.18

hoot

0.17

 cáº£

0.16

REDIENT

0.15

 provisions

0.15

 among

0.15

/un

0.15

afx

0.15

 prominently

0.15

Activations Density 0.056%

phrases that denote inclusion or mention inclusion directly

No Comments

No Known Activations