INDEX

Explanations

phrases related to justice and social issues

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

azel

-0.69

erred

-0.63

VERTISEMENT

-0.62

 explan

-0.61

omore

-0.61

oping

-0.60

 extensively

-0.59

 behavi

-0.59

OIL

-0.59

cius

-0.59

POSITIVE LOGITS

 essence

0.88

same

0.82

liest

0.79

 same

0.79

 embodiment

0.78

 poorest

0.77

 wealthiest

0.76

 richest

0.75

 highest

0.75

 necessities

0.74

Activations Density 0.468%

phrases related to justice and social issues

No Comments

No Known Activations