INDEX

Explanations

instances of military airstrikes and targeted attacks

oai_token-act-pair · gpt-3.5-turbo

terms related to military actions and airstrikes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãĤ´ãĥ³

-0.98

ĸļ

-0.86

 Lifetime

-0.83

vation

-0.73

gebra

-0.72

imensional

-0.72

orial

-0.72

)</

-0.72

zek

-0.71

assic

-0.70

POSITIVE LOGITS

 targets

1.18

 suspected

1.12

 Houth

1.10

 insurgents

1.04

PKK

0.99

 rebel

0.99

 militants

0.97

 strongh

0.97

 ISIL

0.95

 rebels

0.95

Activations Density 0.188%

instances of military airstrikes and targeted attacks

terms related to military actions and airstrikes

No Comments

No Known Activations