INDEX

Explanations

the name "Nicole."

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_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

HAHAHAHA

-0.76

INESS

-0.76

EEK

-0.75

alach

-0.74

chair

-0.73

HAHA

-0.72

ansas

-0.71

lance

-0.71

quartered

-0.71

hips

-0.70

POSITIVE LOGITS

 Simpson

0.92

tti

0.86

 Scher

0.84

tes

0.77

lli

0.76

stal

0.75

tta

0.72

 Cout

0.71

tto

0.71

 Nunes

0.69

Activations Density 0.030%

the name "Nicole."

No Comments

No Known Activations