INDEX

Explanations

words related to political and religious leaders, events, and organizations, particularly those associated with conflicts and human rights issues

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Configuration

neuronpedia/gpt2-small__res_scl-ajt/6-res_scl-ajt

Prompts (Dashboard)

12,288 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

46,080

Data Type

torch.float32

Hook Point

blocks.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

apollo-research/Skylion007-openwebtext-tokenizer-gpt2

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ride

-0.66

 Skydragon

-0.63

esville

-0.61

ilater

-0.61

sburg

-0.60

suit

-0.59

rooms

-0.59

side

-0.59

suits

-0.58

mentation

-0.58

POSITIVE LOGITS

Äĩ

0.70

Å¡

0.68

opsis

0.67

pora

0.66

qua

0.65

cipl

0.65

ocese

0.64

ãĤ´ãĥ³

0.64

agh

0.63

inity

0.62

Activations Density 8.175%

words related to political and religious leaders, events, and organizations, particularly those associated with conflicts and human rights issues

No Comments

No Known Activations