INDEX

Explanations

financial terms related to loss

oai_token-act-pair · gpt-3.5-turbo

references to financial losses

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.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

 tranquil

-0.69

 liking

-0.65

ALLY

-0.64

vez

-0.60

ENTS

-0.60

 nudity

-0.59

REDACTED

-0.59

gently

-0.58

approved

-0.57

opathic

-0.57

POSITIVE LOGITS

 incurred

0.97

 losses

0.97

urance

0.81

 margins

0.78

 earnings

0.77

esville

0.77

igan

0.75

hooting

0.74

igue

0.74

 margin

0.74

Activations Density 0.049%

financial terms related to loss

references to financial losses

No Comments

No Known Activations