INDEX

Explanations

mentions or references to the Pentagon

oai_token-act-pair · gpt-3.5-turbo

references to the Pentagon

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

gger

-0.75

PER

-0.73

âĸ¬

-0.73

âĸ¬âĸ¬

-0.72

Interstitial

-0.71

lihood

-0.71

à

-0.71

cess

-0.69

combe

-0.67

nder

-0.65

POSITIVE LOGITS

 Papers

1.02

 Pentagon

0.96

arium

0.85

selage

0.84

 embassy

0.81

 annex

0.76

 Defense

0.74

 brass

0.74

iets

0.73

eteria

0.73

Activations Density 0.011%

mentions or references to the Pentagon

references to the Pentagon

No Comments

No Known Activations