INDEX

Explanations

specific phrases in a technical or educational context relating to specific entities or concepts

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ĸļ

-0.84

most

-0.82

ħĭ

-0.75

Password

-0.73

ores

-0.72

few

-0.72

rers

-0.71

 Cosponsors

-0.71

uers

-0.71

atu

-0.70

POSITIVE LOGITS

 luck

0.90

 misinformation

0.88

 overlap

0.85

 ammunition

0.78

 firepower

0.78

fun

0.78

 fame

0.78

 information

0.76

 variability

0.76

 caution

0.75

Activations Density 1.445%

specific phrases in a technical or educational context relating to specific entities or concepts

No Comments

No Known Activations