INDEX

Explanations

information related to verification, confirmation, substantiation, and validation of claims, events, and data

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 actionGroup

-0.85

ktop

-0.74

anie

-0.72

atra

-0.69

odan

-0.68

zie

-0.67

ahime

-0.67

ummer

-0.66

 downtime

-0.65

jee

-0.64

POSITIVE LOGITS

 assertions

1.36

 assertion

1.20

 debunked

1.17

 conjecture

1.14

 hypotheses

1.13

 claims

1.12

 factual

1.07

 allegation

1.07

 falsehood

1.06

 validity

1.06

Activations Density 3.388%

information related to verification, confirmation, substantiation, and validation of claims, events, and data

No Comments

No Known Activations