INDEX

Explanations

references to extremist groups, specifically associations with ISIS and its activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.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

 nutrit

-0.74

 ellipt

-0.71

 humid

-0.68

 clocks

-0.67

 toile

-0.67

 leisure

-0.67

ilts

-0.67

erous

-0.67

 postage

-0.66

 toilet

-0.66

POSITIVE LOGITS

 Daesh

1.14

 Taliban

1.12

 ISIL

1.11

 militants

1.11

Nusra

1.10

 Islamist

1.09

 ISIS

1.08

 Islamic

1.06

Islamic

1.05

 Boko

1.04

Activations Density 0.199%

references to extremist groups, specifically associations with ISIS and its activities

No Comments

No Known Activations